152
Mining patterns in genomic and clinical cancer data to characterize novel driver genes Rachel D. Melamed Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy under the Executive Committee of the Graduate School of Arts and Sciences COLUMBIA UNIVERSITY 2015

Rachel D. Melamed

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Rachel D. Melamed

   

Mining patterns in genomic and clinical cancer data to characterize novel driver genes

Rachel D. Melamed

 

                       

Submitted in partial fulfillment of the

requirements for the degree

of Doctor of Philosophy

under  the  Executive  Committee    

of the Graduate School of Arts and Sciences

COLUMBIA UNIVERSITY

2015

Page 2: Rachel D. Melamed

©2015

Rachel D. Melamed

All rights reserved

Page 3: Rachel D. Melamed

ABSTRACT

Mining patterns in genomic and clinical cancer data to characterize novel driver genes

Rachel D. Melamed

Cancer research, like many areas of science, is adapting to a new era characterized by increasing

quantity, quality, and diversity of observational data. An example of the advances, and the

resulting challenges, is represented by The Cancer Genome Atlas, an enormous public effort that

has provided genomic profiles of hundreds of tumors of each of the most common solid cancer

types. Alongside this resource is a host of other data and knowledge, including gene interaction

databases, Mendelian disease causal variants, and electronic health records spanning many

millions of patients. Thus, a current challenge is how best to integrate these data to discover

mechanisms of oncogenesis and cancer progression. Ultimately, this could enable genomics-

based prediction of an individual patient’s outcome and targeted therapies, a goal termed

precision medicine. In this thesis, I develop novel approaches that examine patterns in

populations of cancer patients to identify key genetic changes and suggest likely roles of these

driver genes in the diseases.

In the first section I show how genomics can lead to the identification of driver alterations in

melanoma. The most recurrent genetic mutations are often in important cancer driver genes: in a

newly sequenced melanoma cohort, recurrent inactivating mutations point to an exciting new

melanoma candidate tumor suppressor, FBXW7, with therapeutic implications.

But each tumor is unique, underlining the fact that recurrence will never capture all relevant

mutations responsible for the disease. Tumors are a result of random events that must collaborate

Page 4: Rachel D. Melamed

to endow a cell with all of the invasive and immortal properties of a cancer. Some combinations

of events are lethal to a developing tumor, while other combinations are simply not preferentially

selected. In order to discover these complex patterns, I develop a method based on the joint

entropy of a set of genes, called GAMToC. Using GAMToC, I identify sets of recurrently altered

genes with a strongly non-random joint pattern of co-occurrence and mutual exclusivity. Then, I

extend this method as a means of identifying novel genes with a role in cancer, by virtue of their

non-random pattern of alteration. Insights into the roles of these novel drivers can come from

their most strongly co-selected partners.

In the final section of the main text, I develop the use of cancer comorbidity, or increased cancer

risk, as a novel data source for understanding cancer. The recent availability of clinical records

spanning a large percentage of the American population has enabled discovery of many cancer

comorbidities. Although most cancers arise as a result of somatic mutations accumulating over a

patient’s lifespan, mutations present at birth could predispose some rare populations to increased

cancer risk. Mendelian disease phenotype provides strong insight into the genotype of an afflicted

individual. Thus, if Mendelian diseases with cancer comorbidity can be shown to have specific

defects in processes that are important in the development of that cancer, statistical comorbidity

could provide a new a resource for prioritizing Mendelian disease genes as novel cancer related

genes. For this purpose, I integrate clinical comorbidity, Mendelian disease causal variants, and

somatic genomic profiles of thousands of cancers. I demonstrate that comorbidity indeed is

associated with significant genetic similarity between Mendelian diseases and the cancers these

patients are predisposed to, suggesting highly interesting and plausible new candidate cancer

genes. While cancer may be the result of a series of selected random events, patterns of incidence

across large populations, as measured by genomics or by other phenotypes, contain much non-

random signal yet to be mined.

Page 5: Rachel D. Melamed

TABLE OF CONTENTS

LIST OF GRAPHS, IMAGES, AND ILLUSTRATIONS .............................................................. iv  

LIST OF SUPPLEMENTARY TABLES ....................................................................................... vi  

ACKNOWLEDGEMENTS ........................................................................................................... vii  

1   INTRODUCTION ...................................................................................................................... 1  

2   Coding mutations influencing development of melanoma and nevi .......................................... 9  

2.1   Sequencing melanomas and discovery of FBXW7 as a melanoma tumor suppressor ......... 9  

2.1.1   Methods ...................................................................................................................... 10  

2.1.2   Results ......................................................................................................................... 11  

2.2   Sequencing nevi: exploring the progression to melanoma ................................................ 13  

2.2.1   Methods ...................................................................................................................... 14  

2.2.2   Results and discussion ................................................................................................ 15  

2.3   Discussion .......................................................................................................................... 20  

3   Applying the total correlation to identify and contextualize driver alterations ....................... 22  

3.1   An information theoretic method to identify combinations of genomic alterations that

promote glioblastoma .................................................................................................................. 23  

3.1.1   Introduction ................................................................................................................. 23  

3.1.2   Method ........................................................................................................................ 27  

3.1.3   Results ......................................................................................................................... 33  

3.1.4   Discussion ................................................................................................................... 43  

3.2   GAMToC-L: Using patterns of co-selection of cancer genes to identify and contextualize

novel drivers ................................................................................................................................ 48  

Page 6: Rachel D. Melamed

ii

3.2.1   Methods ...................................................................................................................... 50  

3.2.2   Results ......................................................................................................................... 56  

3.2.3   Discussion ................................................................................................................... 62  

4   Genetic similarity between cancers and comorbid Mendelian diseases identifies candidate

driver genes .................................................................................................................................... 64  

4.1   Introduction ....................................................................................................................... 65  

4.2   Comparing Mendelian disease and comorbid cancer ........................................................ 67  

4.2.1   Integration of disease comorbidities and genes .......................................................... 67  

4.2.2   Genetic similarity of comorbid diseases ..................................................................... 71  

4.3   Mendelian disease comorbidity and cancer processes ...................................................... 78  

4.3.1   Prediction of diseases with shared cellular processes ................................................. 78  

4.3.2   Pan-cancer Mendelian associations ............................................................................ 87  

4.4   Discussion .......................................................................................................................... 91  

5   Data-driven discovery of seasonally linked diseases from an Electronic Health Records

system ............................................................................................................................................. 95  

5.1   Introduction ....................................................................................................................... 96  

5.2   Methods ............................................................................................................................. 99  

5.2.1   Quantifying incidence of diagnoses ............................................................................ 99  

5.2.2   Correcting for confounding trends .............................................................................. 99  

5.2.3   Evaluating periodicity ............................................................................................... 100  

5.2.4   Comorbidity analysis ................................................................................................ 101  

5.3   Results ............................................................................................................................. 102  

5.3.1   LSP-detrend: finding periodic signal ........................................................................ 102  

Page 7: Rachel D. Melamed

iii

5.3.2   Major types of periodic signal and known seasonal disease .................................... 104  

5.3.3   Confirmation of recent reports of seasonal effects ................................................... 105  

5.3.4   Novel findings: acute exacerbations of myasthenia gravis ....................................... 108  

5.3.5   Dissecting causes of seasonality of acute exacerbations by finding comorbid diseases

109  

5.3.6   Comparison between hospital systems ..................................................................... 110  

5.4   Discussion ........................................................................................................................ 111  

6   CONCLUSION ...................................................................................................................... 113  

7   Supplementary Tables ............................................................................................................ 118  

8   REFERENCES ....................................................................................................................... 124  

Page 8: Rachel D. Melamed

iv

LIST OF GRAPHS, IMAGES, AND ILLUSTRATIONS

Figure 1-1: Subtypes of glioblastoma. .............................................................................................. 5  

Figure 2-1: FBXW7 as a novel therapeutic target in melanoma .................................................... 13  

Figure 2-2: Mutation spectrum of nevus types, and frequency thresholds derived per nevus ....... 16  

Figure 3-1: Workflow of GAMToC gene set finding. ................................................................... 26  

Figure 3-2 Illustration of copy number linkages in the GBM cohort ............................................. 28  

Figure 3-3: Visualization of the relationship between temperature (in legend), change in total

correlation (x-axis), and acceptance probability in the simulated annealing (y-axis). ........... 32  

Figure 3-4: Ability to find multi-gene co-mutational patterns. ..................................................... 34  

Figure 3-5: Comparison of different methods in GBM mutation data. .......................................... 36  

Figure 3-6: Recovery of different module sizes in only mutation data. ......................................... 38  

Figure 3-7: Networks of total correlation modules. ....................................................................... 39  

Figure 3-8: Networks seeded with query genes. ............................................................................ 42  

Figure 3-9 Cell cycle, DNA damage, and mitogenic gene subtype associations. .......................... 47  

Figure 3-10 Effect of decreasing temperature ................................................................................ 51  

Figure 3-11 Distribution of the frequency of genes and gene pairs appearing in the module data.

................................................................................................................................................ 53  

Figure 3-12 Frequency of co-selection of pairs of genes in the module data. ................................ 54  

Figure 3-13 GAMToC-L module for the GBM data. ..................................................................... 57  

Figure 3-14: Second module from GBM data. For legend see Figure 3-13. .................................. 60  

Figure 3-15 Module for lower grade glioma. For legend see Figure 3-13. .................................... 62  

Figure 4-1 Distribution of number of genes per disease. ............................................................... 69  

Figure 4-2 Characteristics of Mendelian diseases .......................................................................... 70  

Figure 4-3 Outline of the approach. ............................................................................................... 71  

Page 9: Rachel D. Melamed

v

Figure 4-4 Genes shared in comorbid diseases .............................................................................. 73  

Figure 4-5 Aggregate similarity of comorbid diseases ................................................................... 76  

Figure 4-6 Depiction of comorbid diseases with skin melanoma .................................................. 79  

Figure 4-7 Analysis of the role of albinism related genes in melanoma. ....................................... 81  

Figure 4-8 Pairwise pathway metric for Rubinstein-Taybi and melanoma .................................... 82  

Figure 4-9 Coexpression of ectodermal dysplasia genes with PTK6 ............................................. 83  

Figure 4-10 GSEA plot of the ectodermal dysplasia candidates .................................................... 84  

Figure 4-11 Interaction of Diamond-Blackfan anemia genes with glioblastoma altered genes. .... 85  

Figure 4-12 GSEA plot of holoprosencephaly candidate genes ..................................................... 87  

Figure 4-13 The distribution of the number of comorbid cancer diagnosis codes per Mendelian

disease ..................................................................................................................................... 88  

Figure 4-14 Mendelian diseases with broad cancer links ............................................................... 90  

Figure 5-1: Identifying confounding factors in temporal diagnosis ............................................. 103  

Figure 5-2: Pre-processed and row-normalized monthly incidence for 227 codes with periodic

signal. .................................................................................................................................... 104  

Figure 5-3: Selected diseases with periodic signal. ...................................................................... 107  

Figure 5-4: Overall seasonality of hospitalization in Columbia and Stanford ............................. 110  

Page 10: Rachel D. Melamed

vi

LIST OF SUPPLEMENTARY TABLES Supplementary Table 1: Significantly mutated genes in the melanoma cohort, and their mutations

across the tumors .................................................................................................................. 118  

Supplementary Table 2: Genes significantly less frequently mutated in the nevus cohort, see

2.2.2.4 ................................................................................................................................... 119  

Supplementary Table 3: Pairs of comorbid and genetically similar Mendelian disease and cancer,

related to 4.3. Columns described below: ........................................................................... 120  

Supplementary Table 4: Continuation of Supplementary Table 3 ............................................... 122  

Supplementary Table 5: ADAMS results for comorbidity with acute exacerbations of myasthenia

gravis. Related to section 5.3.5. ............................................................................................ 123  

   

Page 11: Rachel D. Melamed

vii

ACKNOWLEDGEMENTS I would like to acknowledge all who contributed to this dissertation. First, my advisor from my

days before graduate school, Christophe Benoist, who encouraged me to pursue the PhD. I would

like to thank my graduate advisor Raul Rabadan for his support in his lab, as well as all of my

other co-authors on related work. In particular, Andrey Rzhetsky, Jiguang Wang, Antonio

Iavarone, Hossein Khiabanian, Julide Celebi, and Iraz Aydin contributed to work described in this

dissertation. I am grateful to all members of my committee for supervising this work: Raul

Rabadan, Antonio Iavarone, George Hripcsak, Harmen Bussemaker, and Yufeng Shen.

Additionally, I would like to thank members of the Biomedical Informatics department

administration for years of help. On a personal note I thank friends from graduate school for all of

their moral support in the process, including Bo-Juen Chen, Denesy Mancenido, Felix Sanchez

Garcia, Regina Lutz, Francesco Abate and others. My closest friends and loved ones Tommy and

Samantha made it possible for me to continue during some difficult times, and I hope I can do the

same for them.

Page 12: Rachel D. Melamed

1

1 INTRODUCTION

Since the discovery of the first oncogenic genetic lesions, the hunt for causal mutations driving

cancer has been driven by the promise of understanding the basic biology of tumors, forecasting

patient outcome, and finding druggable targets. The Philadelphia chromosome, a chromosomal

fusion discovered via cytogenetic analysis of leukemia patients in 1960 (Nowell and Hungerford

1960), leads to hyperactivation of the Abl tyrosine kinase and sustained, self-sufficient growth

and proliferation signaling. Excitingly, the Philadelphia chromosome is found in 95% of chronic

myelogenous leukemias, and the fusion protein can be inhibited with imatinib, with effective

clinical response(Druker et al. 2001). Others suspected that this growth signaling pathway may

harbor other common mutations that drive tumor growth, and sequenced the BRAF gene in a

number of tumors, finding recurrent activating mutations at a single site, particularly in

melanoma (Davies et al. 2002). This led to development of vemurafinib to target this mutation.

However, vemurafinib is known to fail by 18 months after treatment of melanoma

patients(Poulikakos and Rosen 2011). Additionally, only around 60% of melanomas have BRAF

mutation. Most cancers appear to have far more heterogeneity, and more complexity, than was

initially hoped.

In fact, knowledge about the processes underlying cancer development can suggest explanations

for the heterogeneity between tumors, and for the difficulties in treating some cancers. Cancer

usually requires complementary alterations to multiple cellular functions. For example, the same

BRAF activating mutation found in many melanomas, or a similarly activating mutation of NRAS,

is also found in most benign nevi. The acquisition of this mutation does lead to some

proliferation in these cells, resulting in the nevus, but these growths are benign, self-limited by

the phenomenon of oncogene induced senescence (Michaloglou et al. 2005).

Page 13: Rachel D. Melamed

2

The necessity to bypass this checkpoint represents just one obstacle the tumor must overcome. A

number of changes are necessary in the tumor evolutionary process, including reduced

susceptibility to growth inhibition signals, angiogenesis, ability to invade surrounding tissue, and

genomic instability that can accelerate the evolutionary process (Hanahan and Weinberg 2011).

The necessity for multiple alterations in cancer development was supported decades

ago(Armitage and Doll 1954). Armitage and Doll considered that advanced age of onset for most

cancers may reflect the time it takes for multiple mutations to accumulate in a cell, precipitating

transformation. Modeling the distribution of cancer onset age, they found it was consistent with

the requirement for multiple successive mutations to occur. One rare young-onset cancer is

retinoblastoma, which includes a familial form often involving multiple tumors over a lifetime.

Knudson modeled the variable number of tumors per patient, and age of onset, in familial

retinoblastoma. The analysis suggested a “two-hit” hypothesis, where the germline mutation

present in all cells of all carriers must be accompanied by a somatic mutation(Knudson 2001).

Even in relatively simple cancers with a strong inherited genetic component, random somatic

mutations determine cancer development. Multiple somatic mutations are required for

tumorigenesis in most common cancers.

The findings of recurrent genetic changes, such as BCR-Abl fusion and BRAF mutation, as well as

the understanding that cancer is the result of heterogeneous combinations of somatic alterations,

have encouraged the development of larger scale cancer genomics efforts. The International

Cancer Genome Consortium, and, in the USA, The Cancer Genome Atlas (TCGA), aim to profile

most common cancers. These projects collect complex profiles including mutation, copy number,

gene expression, and epigenetics, all with the goal of understanding how genes are mis-regulated

Page 14: Rachel D. Melamed

3

in cancer. Thus far, TCGA has unrestricted releases of copy number and whole exome

sequencing data for thousands of patients across 19 cancers.

One of the principal goals of somatic genomic tumor collections continues to be to distinguish

“driver” genes from the large number of “passenger” mutations randomly mutated and passively

retained. Drivers include oncogenes like BRAF that promote tumor growth, as well as tumor

suppressors, genes that can inhibit growth and are often disabled in the cell. Prevalent current

methods for identifying drivers look for the most recurrent genetic lesions across cancer

genomics profiles, such as copy number or whole exome sequencing results. Copy number

aberrations (CNA) can indicate deletions of a gene, that can disable tumor suppressor genes by

lowering or abolishing their expression, as well as gene amplifications, which result in extra

copies of a gene and overly expressed, and thus overly active, oncogenes. But CNA inherently

capture background noise: these alterations rarely target a single gene, but may involve large

sections of a chromosome. The most successful algorithm to find significantly altered genes from

CNA is GISTIC, which scores alterations by their recurrence as well as their narrowness(Mermel

et al. 2011). More localized genetic lesions can be found using whole exome sequencing, which

can identify coding mutations that can activate or disable a gene. But much like copy number

data, nucleotide sequence data suffers from many passenger mutations. Studies showed that

recurrent mutations in a gene could be due to processes unrelated to its role as a driver, including

gene length, sequence content, or low constitutive expression (Lawrence et al. 2013). Using

recurrence alone to find driver genes has had much success, but many limitations exist(Lawrence

et al. 2014).

Besides the technical challenge imposed by the prevalence of passenger mutations, another

problem is more basic to the premise that driver genes will be highly recurrently altered across

Page 15: Rachel D. Melamed

4

patients. It is well known that many routes to cancer exist, and this can clearly influence the

prevalence of a driver gene across samples. In glioblastoma, tumors have often been classified as

primary, occurring in older patients, or a less common secondary type, occurring in younger

patients as a progression from a lower grade neoplasm. The two tumor varieties may have no

visible histological distinction, but they have been shown to carry different genetic lesions and

different prognoses. Using gene expression profiles of samples from TCGA glioblastoma project,

Verhaak et al. clustered glioblastomas into four subtypes, connecting the subtypes to different

sets of copy number or point mutations, as well as distinctions in disease phenotype (Verhaak et

al. 2010). For example, the proneural group contains more secondary glioblastomas, has different

treatment response, and may show better prognosis. Subsequent work further characterized

glioblastomas by their methylation profiles, finding a subset of mostly proneural glioblastomas

with a distinct profile of CpG island hypermethylation (G-CIMP subtype)(Noushmehr et al. 2010;

Brennan et al. 2013). These had a distinctive clinical phenotype and profile of copy number

alterations and mutations, as compared to other proneural tumors, as is shown in Figure 1-1. The

authors suggest that these genetic alterations cooperate specifically with the methylation-induced

gene silencing in these tumors.

Page 16: Rachel D. Melamed

5

Figure  1-­‐1:  Subtypes  of  glioblastoma.    

As  clustered  by  methylation  profiles  (“DM  CLUSTER”),  clusters  show  differences  in  clinical  indicators  as  well  as  gene  expression  clusters  (“EXP  CLUSTER”)  and  presence  of  somatic  copy  number  aberrations  or  mutations.    Reproduced  from  (Brennan  et  al.  2013)  figure  5.  

This type of analysis shows that context is important in understanding driver mutations, and less

frequent subtypes, such as G-CIMP tumors, may have their own sets of highly relevant drivers.

Melanoma provides other examples of the heterogeneity among tumors. A variety of alterations

occur to activate the pro-growth MAPK pathway: BRAF gain of function mutations occur in over

50% of patients, with NRAS mutations in another 20% or more (Watson et al. 2013). But these

two members of the MAPK pro-growth pathway almost never co-occur, either because of lack of

selective advantage to further disruption of the MAPK pathway, or because such co-mutation

proves deleterious. This type of pattern has led to the “exclusivity hypothesis”, which states that

redundant mutations are less likely to be found in the same tumor. Thus, mutually exclusive

mutations could be informative of functional relationships between genes. Additionally, the

mutual exclusivity and prevalence of these two alterations suggests that activation of the MAPK

pathway may be crucial for melanoma development. Intriguingly, some melanoma subtypes

harbor no BRAF or NRAS mutations, including acral and mucosal melanomas that often bear

activating KIT mutations that may influence the same pathway. It is important to note that no two

Page 17: Rachel D. Melamed

6

changes are truly redundant: even NRAS and BRAF, which are adjacent in the signaling network,

have different functions. While both NRAS and BRAF mutations activate MAPK, NRAS

additionally activates the PI3K pathway(Palmieri et al. 2009). In conclusion, much like in

glioblastoma, melanoma shows a variety of pathways to tumor development, and shows how

cancer alterations can be informative of each other.

To summarize current work in finding relevant alterations in cancer genomes, much emphasis has

been placed on finding the most recurrent changes. But simultaneously, it is understood that many

pathways to cancer exist, including subtypes of tumors. Mutations that are less prevalent across

the population can still be highly relevant for an individual tumor. Understanding subtypes of

cancer is as important a question as finding driver genes, and the two goals are highly interlinked.

Finding subtypes of cancer can help understand pathways to the disease: the subtypes of

glioblastoma have been linked to different neural cell types. Understanding the commonalities

and distinctions among sets of tumors will provide a more complete picture of tumor biology.

My goal in this dissertation is to find important genes genetically altered in cancer, but also to

understand how these lead to tumor development. I apply current tools to find the recurrent genes

of interest across compilations of tumor samples, but my main focus is in developing new

approaches to using large cancer genomics compendia to understand cancer biology. In chapter 2

I describe work characterizing genetic alterations driving melanoma, the most lethal skin cancer,

and a cancer that is incurable in its metastatic form. Due to high rates of ultraviolet induced DNA

damage, melanoma genomic profiles are highly complex, with hundreds of protein changing

mutations per patient. In newly sequenced cases of melanoma, I work on identifying genes with

evidence of selective alteration in the disease. As nevi are a risk factor for melanoma, and

Page 18: Rachel D. Melamed

7

dysplastic nevi are considered to be possible melanoma precursors, I also compare dysplastic and

other nevi to melanoma to better understand the progression from a benign nevus to a melanoma.

Beyond current methods for finding recurrent genes driving oncogenesis, I develop novel

approaches to identify important genes. I propose in this work that in genomic and clinical

profiles of populations of cancer patients contain patterns that are an underutilized source of

knowledge about cancer biology. The approaches I develop mine these patterns for new evidence

of genetic alterations driving cancer development, with a particular focus on melanoma and

glioblastoma. In section 3, I develop methods using measures from information theory to find

mutation patterns. In 3.1 I describe an algorithm for finding Genetic Alteration Modules with

Total Correlation, or GAMToC. This method addresses the combinatorial nature of genetic

alterations in cancer. The examples above provide ample context for the concept: a certain

combination of genetic lesions are present in subtypes of glioblastoma, and MAPK activating

mutations are mutually exclusive across cases of melanomas. GAMToC is an information

theoretic approach to find these patterns across compilations of cancers, and to exploit these types

of patterns to find driver genes and to understand their role in cancer development. I show results

of the method in glioblastoma, and, motivated by these observations, I extend this method in 3.2,

which describes GAMToC-L. While the first version of GAMToC searched for combinations of

highly recurrent genetic alterations across tumor compendia, GAMToC-L uses all genomic

information. I consider that a non-random pattern of joint genetic mutation that includes a gene

can help us discover new less recurrent drivers in caner.

My work with GAMToC developed an unsupervised method that relies only on patterns of

mutation within a collection of cancer samples. In the final section I will bring together multiple

sources of information to find driver genes. In this chapter, I discuss the new possibility of using

Page 19: Rachel D. Melamed

8

Electronic Health Records (EHR) to find novel cancer-related genes. Included in supplementary

chapter 5 is a brief discussion of the utility of EHR for finding patterns of disease. While

GAMToC looked for combinations of genetic alterations, the approach described in chapter 4

uses combinations of co-occurring Mendelian disease and cancer. As the mutations that are

responsible for the Mendelian disease are usually known, I present clinical co-occurrence as a

novel source for identifying cancer drivers.

 Genomic profiles of tumors, including genetic mutations, gene expression, and epigenetics, can

exquisitely characterize a particular tumor in a particular patient. Francis Collins, director of the

National Institutes of Health, has called such high dimensional genomic data “the leading edge of

precision medicine.” But precision medicine requires advances in basic science that can identify

the pathways that are most relevant in cancer development, and thus the mutations and other

cellular changes that likely drive a particular patient’s tumor. Novel approaches to identifying

patterns in large datasets will indicate the selective processes shaping cancer, and common ways

that tumors overcome these obstacles. Using large patient cohorts, we can thus better understand

how the disease arises in each patient, and we can identify vulnerabilities that can be exploited in

targeted therapies. This thesis illustrates some of the challenges, solutions, and opportunities

created by the technological revolutions in data acquisition and high throughput genomic

technology.

 

Page 20: Rachel D. Melamed

9

2 Coding mutations influencing development of melanoma and nevi

Metastatic melanoma is a highly lethal disease with no effective current treatment. Risk factors

include chronic exposure to mutagenic ultraviolet light and the presence of dysplastic nevi,

suggesting a progressive evolutionary process leading to the cancer. However, melanoma can also

arise from sites lacking sun exposure, and some distinctions in mutation profiles have been

associated with both body location and sun exposure. Despite the highly recurrent presence of

mutations in BRAF and NRAS, and the understanding of many pathways involved in the disease,

the high background mutation rate in melanoma makes identification of driver genes difficult. It

is a heterogenous tumor with a highly complex landscape of genomic alterations. Here, we report

two approaches to discovery of driver genes in melanoma. First, we perform whole exome

sequencing on a cohort of melanomas. We examine the resulting mutation profiles for recurrent

alterations, and then we further investigate for genes with evidence of potential as a therapeutic

target. The results from this section are published in (Aydin et al. 2014). Second, we sequence a

set of nevi, including dysplastic nevi and congenital melanocytic nevi, to find genes that drive

nevus development, and genes that distinguish a nevus from a melanoma.

   2.1 Sequencing melanomas and discovery of FBXW7 as a melanoma

tumor suppressor

In order to find novel driver genes in melanoma, we perform whole exome sequencing of a small

exploration set of metastatic melanomas. We identify a gene, FBXW7, with evidence for a role as

a novel tumor suppressor in melanoma. FBXW7 has known interaction with the oncoprotein

NOTCH1. Our results from functional validation and in vivo studies suggest that inactivating

mutations in FBXW7 have relevance in indicating notch inhibitors as a treatment for melanoma.

Page 21: Rachel D. Melamed

10

2.1.1 Methods  We screened a cohort of eight metastatic melanomas using whole exome sequencing. Sequencing

resulted in an average of 42 million reads per sample (32 to 101 million), of which an average of

98.4% mapped to the hg19 genome using BWA (H. Li and Durbin 2009), followed by GATK

indel realignment, resulting in an average depth of 11 reads per base covered at depth greater than

zero. Using the SAVI algorithm (Trifonov et al. 2013), we called positions with nucleotide

mutations. From these, we retained only the variants at positions with depth greater than 10 in

both tumor and normal samples, and we filtered out any variants that also appeared in any normal

sample in a greater than 25% of reads. We identified of a total of 2308 exonic mutations, with

737 synonymous and 1571 non-synonymous, consisting of 1431 missense and 78 non-sense

mutations, and 62 insertions/deletions. The mean exonic non-synonymous mutation rate was

10.6 mutations per megabase, with mutation rates varying from 2.8 to 26.7. All cases sequenced

were cutaneous melanomas on sun-exposed sites, and as expected the majority of nucleotide

substitutions were C>T or G>A transitions (73-91% of all mutations), indicative of ultraviolet-

induced damage as well as cytosine deamination. The hot spot mutation, BRAF V600E, was

present in six of the eight cases.

Following sequencing and variant calling, we used the collection of mutations to identify genes

with evidence of positive selection for nonsynonymous mutations. First, we evaluated whether a

gene had more nonsynonymous mutations than would be expected. We estimated the expected

number of nonsynonymous mutations for each gene using the number of synonymous mutations

in that gene, NS,G and the nonsynonymous to synonymous mutation ratio across all genes, NN/NS

resulting in an expected number of nonsynonymous mutations of NS,G*NN/NS. Then we evaluated

whether the observed number of nonsynonymous mutations, NN,G was significantly more than this

expected value using a Poisson model. We also tested for elevated number of mutations given

Page 22: Rachel D. Melamed

11

gene length, using the background number of mutations per coding positions. For this test, we use

a binomial model, with amino acid length corresponding to number of trials, and probability of a

nonsynonymous mutation calculated from the average number of nonsynonymous mutations per

amino acid, across all genes with mutations. Finally, our candidate list contains 23 putative driver

genes that have p < .05 in both tests (Supplementary Table 1).

As we were particularly interested in finding therapeutically related mutations in melanoma, we

searched for genes that might be impacted by our mutations, and that are druggable targets. To

find interacting partners for the mutated genes, we use GeneRIF interactions(Brown et al. 2005),

keeping only those interactions that are documented in human cells and are not based on affinity

assays. Then, for each of these interacting partners, we use the Cancer Commons(Shrager,

Tenenbaum, and Travers 2010) drug target database to assess if partners are druggable. Only a

few recurrently mutated genes, including FBXW7 have druggable interacting partners.

2.1.2 Results Using exome sequencing we first call variants and then select the genes with evidence of positive

selection for their alteration. Then, we use external sources of evidence about gene interacting

partners to identify FBXW7 as a putative melanoma driver of interest. We further support a role

for FBXW7 by sequencing a wider panel of 103 melanomas including 77 tumor samples and 26

cell lines. We sequence the coding regions of FBXW7, BRAF, and NRAS. Non-synonymous

mutations in FBXW7 appear in eight cases (8% frequency), with five nonsense, two missense, and

one frameshift mutation. This is a significantly elevated mutation rate: the probability of having

this number of nonsynonymous mutations is less than 10-4, given the length of the gene (710

amino acids) and the nonsynonymous mutation rate per base in our samples (1 x 10-5). Mutations

within the WD40 domain of FBXW7 are predicted to disrupt substrate binding, and thus lead to

sustained activation of its substrate oncoproteins. Of note, the presence of mutations in FBXW7

Page 23: Rachel D. Melamed

12

does not correlate with BRAF or NRAS mutation status. Collectively, these findings identify

somatic mutations of FBXW7 as a novel recurrent genetic event in melanoma.

 In work to characterize FBXW7’s disruption in the disease, we profile expression of the gene in a

panel of melanomas as compared to benign nevi, and we find that FBXW7 is downregulated in

melanoma, and underexpression correlates with mutation. As FBXW7 has a demonstrated role in

other tumor types, particularly in its interaction with known oncoproteins(Oberg et al. 2001;

Welcker et al. 2004), this gene is of high interest.

Then, to investigate the mechanism of FBXW7’s influence on melanoma, we examine the effect

of FBXW7 loss on known regulated proteins including NOTCH1, its direct target, and other

targets including CCNE1 and MYC. Of these, NOTCH1 is consistently upregulated in cell lines

with loss of FBXW7. As well, we transplanted immunodeficient mice with NRAS mutant

melanocytes, and then used an shRNA targeting FBXW7 to silence its expression. In this

xenograft experimental model, not only is NOTCH1 significantly upregulated, but tumors grow at

accelerated rates compared to a control shRNA (Figure 2-1). Ectopic expression of the mutant

FBXW7 in the NRAS mutant melanocytes also accelerates tumorigenesis. As NOTCH1 appears to

be strongly regulated by FBXW7, and this change appears to influence tumor growth, we create a

set of xenografts bearing the NRAS mutation and FBXW7 knockdown, and we treat the resulting

tumors with a notch inhibitor, dibenzazipine. These tumors show significant reduction in growth

as compared to a control group. Thus, this study suggests a mechanism for activation of

NOTCH1 in melanoma, via the newly identified melanoma tumor suppressor FBXW7, and

suggests that in some melanomas with deregulated NOTCH1 expression, as via FBXW7 ablation,

notch inhibitors may be a useful therapy.

Page 24: Rachel D. Melamed

13

 Figure  2-­‐1:  FBXW7  as  a  novel  therapeutic  target  in  melanoma  

These  figures  are  reproduced  from  (Aydin  et  al.  2014).    A.  mutants  of  FBXW7  are  associated  with  larger  tumor  volume.    B.  mutants  of  FBXW7  are  associated  with  greater  expression  of  its  ubiquitinated  target  NOTCH1,  and  with  upregulation  of  NOTCH1’s  target  HEY1.    C.  Treatment  with  Dibenzazipine  (DBZ)  significantly  reduces  tumor  volume.  

2.2 Sequencing nevi: exploring the progression to melanoma

To further explore factors influencing development of melanoma, we explore the genomic

landscape of nevi and dysplastic nevi. Dysplastic nevi are benign neoplasms of melanocytes that

are considered both risk factors for and possible precursors to melanoma. Patients with a

dysplastic nevus have twofold risk for melanoma, while patients who bear more than ten

dysplastic nevi have a 12-fold increased risk(Elder 2010). Although many melanomas arise de

novo, about 25-50% of melanomas have a histologically associated nevus. As melanoma is

thought to result from progressive alterations, we wished to characterize the genomic landscape

of nevi. The goal is to identify what genetic events separate a benign from malignant state, in

related tissue types. Our panel includes multiple nevi per patient for some patients, as well as

matched normal blood samples for each patient. In this project, we confirm the widespread

presence of the known nevus driver mutations affecting NRAS and BRAF. We also show that nevi

from the same patient display a branched pattern of evolution. Finally, we identify genetic

changes that are distinctly significantly less likely to occur in nevi as opposed to melanoma,

pointing out sets of events the precipitate malignant transformation.

C

Page 25: Rachel D. Melamed

14

2.2.1 Methods 2.2.1.1 Calling significant variants in nevi  We collect 35 nevi from 22 patients, including eight common acquired nevi (CAN), four

congenital melanocytic nevi (CMN), and 23 dysplastic nevi (DNS) from patients with dysplastic

nevus syndrome. We tailor a variant calling strategy to the unique nature of this data: obtaining

pure sample of melanocytes from a nevus is impossible, and purity varies between nevi. Due to

the predicted impurity of the samples, we first examine all nevi for the presence of BRAF V600

and NRAS Q61 mutations. These mutations are present in a variable fraction of reads aligning to

their respective genes, as low as one read, but in multiple reads for 32 of 35 nevi. Thus, we

employ a strategy suited to identifying mutations supported by a low percentage of reads. First,

conservative alignment to the reference genome is used. BWA (H. Li and Durbin 2009) with no

Smith-Waterman mate rescue is followed by realignment of reads that may have been misaligned

due to insertions or deletions. Next, presence of variants is called using the SAVI statistical

procedure (Trifonov et al. 2013). Variants are filtered by the SAVI statistic, strict absence in the

matched blood sample, as well as using read depth, and the presence of supporting reads aligning

to both strands. Another filter uses a sample-specific threshold on the frequency of variant-calling

reads among the reads aligned to the variant position. This sample-specific threshold is

calculated by using the BRAF V600E (or NRAS Q61K/L) variant read depth in that sample as

approximation of the expected heterozygous variant presence per sample. From this presence a

lower bound on expected heterozygous variant frequency is determined using a binomial to

model the read distribution. A minimum frequency of 3% variant is imposed. Finally, presence

of each variant is checked using SamTools (http://samtools.sourceforge.net/SAM1.pdf) quality

greater than one, which takes into account other characteristics of reads that identify a position as

a true variant.

2.2.1.2 Comparison of nevus mutation to melanoma mutation  

Page 26: Rachel D. Melamed

15

We download level 2 somatic mutations from TCGA, as well as the results of the Broad Institute

MutSigCV pipeline. We use MutSigCV’s statistic(Lawrence et al. 2013) to identify those genes

with a significantly recurrent mutation pattern in melanoma. Then, we test whether each gene is

significantly more frequently mutated in the melanoma cohort as opposed to the nevus cohort.

This can be tested using a binomial model of nonsynonymous mutation frequency per nevus or

melanoma. The difference between the binomials could be calculated as a chi-square statistic, or

similarly, using the hypergeometric statistic to test whether nevi are significantly depleted for a

given mutation. This quantifies whether the melanoma genes have a similar, or lesser, rate of

mutation in nevi, as compared to the gene’s mutation rate in melanoma.

2.2.2 Results and discussion 2.2.2.1 Spectrum of mutations in nevi We collect 35 nevi from 22 patients, including eight common acquired nevi (CAN), four

congenital melanocytic nevi (CMN), and 23 dysplastic nevi (DNS) from patients with dysplastic

nevus syndrome. As described in the Method, we combine a liberal sample-specific threshold for

identifying mutations with stringent quality filters that remove many of the variants most likely to

be false positives.

There are a median of 16 nonsynonymous and 9.5 synonymous mutations per nevus, but mutation

profile varies widely, ranging from four nevi with no nonsynonymous mutations to a nevus with

61 nonsynonymous mutations. Broken down by subtypes, two of the four CMN have no

mutations. The other two have five and four mutations, making this type of nevus the least

mutated. Both CAN and DNS have higher mutation rates. CAN have a median of 10

nonsynonymous mutations, with all cases displaying one or more mutations. Mutations rates in

DNS are higher, as might be expected, with a median of 21 mutations per case. However, two of

the DNS nevi have no called mutations, possibly due to low purity. For the dysplastic nevi, CàT

mutations are predominant, consistent with UV induced damage and cytosine demination

Page 27: Rachel D. Melamed

16

mechanisms. Common acquired nevi largely display a similar pattern, while mutations in

congenital nevi, although rare, do not appear to share this pattern (Figure 2-2).

 

Figure  2-­‐2:  Mutation  spectrum  of  nevus  types,  and  frequency  thresholds  derived  per  nevus  

Frac

tion

of m

utat

ions

N

umbe

r of m

utat

ions

DNS CAN CMN

30

20

10

0

Thre

shol

d

Page 28: Rachel D. Melamed

17

To identify potential novel genes beyond BRAF and NRAS that could contribute to nevi, we

compile a list of genes with recurrent alterations. 22 genes are found mutated in two or more nevi.

The list includes long genes known to be commonly somatically altered (e.g. SYNE1, DNAH5),

due to unknown mechanisms. The other genes that present higher mutation rate are: NRAS,

BRAF, NOL4, TEK, PCDHB14 and PCDH15. Interestingly, NOL4 is epigenetically silenced in

squamous cell carcinomas of the head and neck, making this a potential tumor suppressor in these

cancers(Demokan et al. 2014). TEK is a protein tyrosine kinase that is most associated with

endothelial cell growth signaling and vascular development. Both PCDHB14 and PCDH15 are

members of the protocadherin family that are most expressed in neural cell junctions.

2.2.2.2 BRAF or NRAS may be mutated in all nevi Because BRAF and NRAS are known to have activating mutations in melanoma as well as in

dysplastic and congenital nevi, we first examine the presence of these activating mutations in our

collection. We find statistically significant presence of BRAF V600E mutation in 14 nevi, and

NRAS Q61 mutation in four nevi, including two Q61K and two Q61R mutations. These

mutations are present a wide range of frequencies, from 17% to 58% of reads covering these

regions of the exome. This wide range is to be expected given the impurity of the melanocyte

content in the nevus sample, when combined with possible subclonal mutation load and

sequencing error.

We examine how BRAF and NRAS mutation are associated with subtype, and whether we can

rule out mutation to these loci in the 17 nevi with no BRAF or NRAS mutation called. Thus, we

check the reads aligning to the two mutation regions for any presence of these mutations. High

quality sequencing reads supporting the relevant mutations are present at very low frequency in

all remaining nevi, in greater than 4% of the reads in most cases. For the four CMN, two have

Page 29: Rachel D. Melamed

18

called NRAS Q61K/R mutation and the other two have evidence for mutated NRAS at low

frequency. The CAN all have strong BRAF V600E mutation.

The DNS set contains multiple samples of low purity, complicating efforts to ascertain mutation

rates. Of the 25 DNS, six have strong BRAF V600E mutation and two have strong NRAS

Q61K/R mutation. Twelve more of the DNS have BRAF variants called at high quality but in a

low fraction of reads. One more dysplastic nevus has credible evidence of NRAS Q61K mutation,

again at low frequency. The remaining two dysplastic nevi display possible subclonal

composition based on BRAF and NRAS loci: patient P025-nevus-1 has reads supporting both

NRAS Q61K and BRAF V600E mutations, though the NRAS reads are at very low frequency.

Finally, patient P016 has a very interesting pattern of mutation: like the others, this nevus has low

frequency chr7:140453136 AàT mutations, present in four reads, which would confer V600E

mutation. However, two of those reads also have chr7:140453137 CàA mutations, which could

reflect a subclone expressing BRAF V600D. As a comparison, we test the frequency of

mutations to nearby codons, BRAF G604 and NRAS D65. No mutations are found.

The results support mutual exclusivity between BRAF and NRAS within a nevus. Using the more

liberal thresholds for variant calling, we find that only one of the 35 nevi has both BRAF and

NRAS hotspot mutations of high quality and at a greater than 1% frequency (p-value for mutual

exclusivity is 4.9x10-7).

2.2.2.3 Evolution and recurrent mutations in nevi In a subset of patients, multiple nevi are sequenced. This includes patient 8, who presented with

both a congenital nevus and a dysplastic nevus, and patient 6, a classic case of dysplastic nevus

syndrome. We compare nevi from the same patient to each other to investigate two hypotheses

about their pattern of evolution. First, we assess whether the nevi could have any common

Page 30: Rachel D. Melamed

19

precursor cell subsequent to their shared germline. Second, we consider whether the shared

genetic background of nevi has an influence in the sets of mutated genes that result in this

abnormal state.

In order to discover whether nevi share a common post-germline ancestor, we look for identical

mutations in the nevi that are absent in the matched blood sample. As identical BRAF and NRAS

mutations are frequent among all nevi, these are excluded. Another identical mutation to

ASPHD1 is most likely due to germline variants or sequence alignment problems. Thus, we

conclude that different nevi in the same patient evolve in a branched evolutionary pattern.

Next, we look for overlap in the sets of genes mutated in nevi from the same patient, as an

indicator of a shared route to nevus development. We find that two CAN from patient P019 both

have mutations to different positions of NOL4. As the nevi have 13 and 15 mutations in all, they

are highly unlikely to share mutations to this gene by chance. Similarly, two DNS from patient

P6 have mutations to TEK, while the nevi have only 2 and 14 mutations each. These findings

support the hypothesis that convergent evolution occurs in nevi, similar to cancer, and

additionally put forward these two genes as being of particular interest in the development of

nevi.

2.2.2.4 Melanoma-specific genomic alterations As the intention of this study is to better understand the development of melanoma, and the

necessary mutations for a nevus to progress to a melanoma, we compare prevalence of the

mutations in nevi to that in melanoma. First, we extract a list of genes with evidence of positive

selection for nonsynonymous mutation, using a compilation of whole exome sequencing results

from 297 melanoma cases from TCGA, and we estimate the frequency of altered cases from the

297 samples. We statistically determine if nevi have a significantly lower rate of mutation to any

Page 31: Rachel D. Melamed

20

genes, identifying genes that could be the drivers in the transformation from the benign precursor

nevus to the cancer. The full list is shown in Supplementary Table 2. We find that TP53,

CDKN2A, and NF1 are less mutated in the nevi. A number of other interesting genes are

highlighted, including the TP53 related gene TP63, BCLAF1, another apoptosis related gene.

Other genes on the list are related to epithelial cell junction, including DSG3 and COL3A1.

Our results suggest that key changes in the progression of melanoma to nevus include changes

related to the apoptosis response pathway (CDKN2A, TP53, TP63, BCLAF1) and the dermal-

epidermal junction (DSG3, COL3A1). These results can be applied in a therapeutic setting aimed

at understanding if an abnormal tissue sample is a nevus, or a potential melanoma requiring a

more aggressive intervention. Future efforts will help us understand what lesions drive the

transition from nevus to aggressive melanoma.

2.3 Discussion

Melanoma is both a clinically and genetically complex disease, and in its metastatic form it is

incurable. These studies have made steps toward untangling the genomics of this cancer. In a

clinical context, a targeted sequencing panel including FBXW7 could eventually influence

treatment decisions such as use of a notch inhibitor. For patients with dysplastic nevi, targeted

sequencing could ascertain whether a nevus displays any of the mutations that are associated with

melanoma, suggesting a benefit from more aggressive early treatment. However, many more

questions could arise as a result of our findings. It would be interesting to know whether

FBXW7-mutated patients display any particular clinical phenotype, or if they are less likely to

have amplifications of notch genes. This pattern would be expected if FBXW7 inactivation is

sufficient for oncogenic activation of notch. It would also be highly interesting, though

practically difficult, to sequence a nevus before and after transition to melanoma. We do not

know if any of the mutations in nevi, other than the BRAF and NRAS mutations, are propagated in

Page 32: Rachel D. Melamed

21

melanomas. While these small studies allow exploration of specific biological phenomena, with

larger cohorts we can find more complex and informative patterns in cancer data.

Page 33: Rachel D. Melamed

22

3 Applying the total correlation to identify and contextualize driver alterations

While my work in melanoma ranks important cancer genes by mutational recurrence across

compendiums of tumor samples, landscape mutational approaches that score each gene

individually ignore the known effects of mutational context on selection and do not address the

combinatorial complexity of genomic alterations in tumors. Tumors are the result of accumulated

genomic alterations that cooperate synergistically to produce uncontrollable cell growth.

Identifying recurrent alterations among large collections of tumors is only one way to pinpoint

genes that endow a selective advantage in oncogenesis and progression. In my dissertation work I

have intended to go beyond recurrence to find how combinations of genetic changes influence the

development of cancer. In this section, I develop an information theoretic framework that

integrates copy number and mutation data to identify gene modules with any non-random pattern

of joint alteration. A non-random pattern of co-mutated genes is evidence for selective forces

acting on tumor cells that harbor combinations of these genetic alterations. Although existing

methods have successfully identified mutually exclusive gene sets, no current method can

systematically discover more general genetic relationships with no prior knowledge. I develop a

framework and methods termed Genomic Alteration Modules using Total Correlation

(GAMToC), to find combinations of recurrently altered genes with a related pattern of mutation.

Additionally, I present the Seed-GAMToC procedure, which uncovers the mutational context of

any putative cancer gene. All software is publicly available. I apply GAMToC to glioblastoma

multiforme samples, and the results show distinct subsets of co-occurring mutations, suggesting

distinct mutational routes to cancer and providing new insight into mutations associated with

Proneural, Proneural/G-CIMP, and Classical types of the disease. Indeed, considering

combinations of genetic mutations in cancer is a powerful approach to learning about the disease.

Page 34: Rachel D. Melamed

23

This work is under review at the Journal of Molecular Cell Biology. Then, I describe, a follow-up

work, in preparation for submission. This works uses the same principles as in GAMToC, but

instead this approach enables us to find driver mutations in cancer by their pattern of related

mutations. Finding driver alterations in copy number data has proved highly challenging, and the

results suggest new sources of evidence for selected alterations in cancer.

3.1 An information theoretic method to identify combinations of

genomic alterations that promote glioblastoma

3.1.1 Introduction Tumors are known to evolve by acquiring genetic lesions. Each mutation creates a cellular state

uniquely predisposed to thrive with the addition of further specific survival abilities (Hanahan

and Weinberg 2011). Recent studies have successfully exploited the selective pressures on

developing tumors to rank important cancer genes by mutational recurrence across compendiums

of tumor samples (Beroukhim et al. 2007; Mermel et al. 2011; Lawrence et al. 2013). But

approaches that score each gene individually ignore the known effects of mutational context on

selection. Tumor survival can be promoted by damage to only one of a set of alternate genes in a

pathway (mutual exclusivity of aberration), while other genetic changes only provide a selective

advantage to a cancer in a given mutational context (co-occurrence of aberration). For example,

in melanoma, BRAF gain of function mutations occur in 40% of patients and NRAS mutations in

25%, but these two members of the MAPK pro-growth pathway almost never co-occur, either

because of lack of selective advantage to further disruption of the MAPK pathway, or because

such co-mutation proves deleterious (Davies et al. 2002). Despite their frequency, MAPK-

activating mutations alone are an evolutionary dead end for the cancer, resulting in cell

senescence(Michaloglou et al. 2005). Cancer progression also requires disruption of a tumor

suppressor function such as CDKN2A(Michaloglou et al. 2005). This example shows that

complex patterns of mutual exclusivity and co-occurrence of mutation, thus far identified in a

Page 35: Rachel D. Melamed

24

piecemeal fashion, are to be expected across cancer cases. Additionally, the observed mutational

relationships of genes, and thus the context in which a genetic aberration is of benefit to tumor

development, can provide insight into the functions of genes that are altered in cancer.

However, most approaches seeking relationships between cancer mutation events focus on

mutually exclusive lesions, reasoning that this pattern may reflect underlying pathways(Mark D.

M. Leiserson et al. 2013; Vandin, Upfal, and Raphael 2012; Szczurek and Beerenwinkel 2014; C.

A. Miller et al. 2011). But these methods will miss other relationships between mutations, such

as co-occurrence. Additionally, the assumption that different genes in the same pathway are

interchangeable is a strong claim. Combinations of genes have been found to jointly predict

cancer phenotype (Varadan and Anastassiou 2006; Mo et al. 2013), but, to our knowledge, no

unsupervised method exists for finding related genetic alterations.

A different approach has been to scan for representation of dysregulated genes within gene sets

known to be functionally related. Recent studies have found pathways predicted to be perturbed

by differential gene expression (Tarca et al. 2009), or mutation (Boca et al. 2010), or when

multiple sources of information on gene activity are integrated (Vaske et al. 2010). Other

methods have used graph topology to find functional interaction sub-networks enriched in

mutated genes(Vandin, Upfal, and Raphael 2011; Cerami et al. 2010; Hofree et al. 2013; G. Wu,

Feng, and Stein 2010), or to identify cliques of genes with mutually exclusive mutational

occurrence(Ciriello et al. 2011). These approaches have the advantage of being able to use

diverse genome-wide alteration information and provide a biological context for the patterns

discovered, but they rely on known gene interactions and on narrow definitions of gene

interaction.

Page 36: Rachel D. Melamed

25

We propose a method that integrates copy number and point mutation information, does not

require prior functional information, and can find any structured module of genes, rather than

only mutually exclusive alterations. The method, Genomic Alteration Modules using Total

Correlation (GAMToC), selects a gene set with high total correlation. Total correlation measures

the difference between the joint uncertainty, or entropy, of a set of variables (genes), as compared

to their individual uncertainties(Watanabe 1960). When there is no joint relationship between the

variables, the difference will vanish. On the other hand, a high total correlation suggests a joint

relationship among the variables, which is not necessarily linear. Because our method can detect

any sort of dependency between the variables, it is sensitive to unexpected varieties of gene

interactions. It does not require the assumption that different alterations to the same pathway are

more or less interchangeable, and it is not restricted to finding genes only in the same pathway.

Instead, the genomic data can lead us to the combination of functional changes that are

cooperating in the cancer. We present two implementations of GAMToC, one that uses a greedy

method to find a single module starting from a pair of related genes, and another that uses a

Simulated Annealing (SA) method to find the highest-scoring gene set. We examine the speed of

the two implementations as compared to exhaustive search, and we evaluate their sensitivity in

simulated data. Then, we apply the method to glioblastoma multiforme (GBM) copy number and

mutation data from the TCGA. Additionally, in Seed-GAMToC we make use of the same

principles to characterize query genes with a likely, but unclear, role in cancer progression by

finding a module that contains genes with a related pattern of selection.

We apply GAMToC to copy number and nucleotide mutation measurements from The Cancer

Genome Atlas (TCGA) glioblastoma project (“Comprehensive Genomic Characterization Defines

Human Glioblastoma Genes and Core Pathways.” 2008), as summarized in Figure 3-1. We are

able to recapitulate known gene interactions, and we additionally recover genes associated with

Page 37: Rachel D. Melamed

26

subtypes of glioblastoma. Our results suggest that specific alterations to key cancer pathways are

not equivalent: on the contrary, there are clear contexts where functionally related genes are

differentially selected for alteration. Thus our method is uniquely suited to find and characterize

genes that are related in cancer development. The software is freely downloadable and can be

applied to any copy number and point mutation data set.

Figure  3-­‐1: Workflow of GAMToC gene set finding.

Genomic alterations (e.g., CNAs and somatic mutations) are integrated to create a binary matrix of samples and genes. The total correlation score compares the entropy of the mutational statuses of individual genes (labeled 1 through 4) against their joint entropy, in effect testing the hypothesis that these gene mutational statuses have a relationship (indicated by the connected network). GAMToC finds sets of mutationally related genes using this score, and we visualize the results in a pairwise correlation network.

Binary matrix describing patients and mutated genes

g1# g2#

g3#g4#

g1# g2#

g3#g4#

Greedy method, simulated annealing, and visualization of GAMToc Module as pairwise network

Somatic Mutation

Somatic Copy Number

Alteration

Tumors

Gen

es

t1 t2 t3 …g1g2g3g4…

t1 t2 t3 …g1g2g3g4…

t1 t2 t3 …g1g2g3g4…

t1 t2 t3 …g1g2g3g4…

t1 t2 t3 …g1g2g3g4…

t1 t2 t3 …g1g2g3g4…

TC(g1,g1,!,gn ) = H (gi )i=1

n

∑ −H (g1,g2,!,gn )

H (g1,g2,!,gn )

H (gi )i=1

n

Page 38: Rachel D. Melamed

27

3.1.2 Method 3.1.2.1 Preprocessing genetic aberration data Currently the GAMToC algorithm can start from assessments of sample copy number aberrations

and from nucleotide variant calls resulting from whole exome sequencing (WES) data. For the

TCGA GBM data, we downloaded processed data from the Broad Institute Firehose

(http://www.broadinstitute.org/cancer/cga/Firehose) download data set of 9/23/2013. This

includes mutation calls, GISTIC2 results, and thresholded calls of copy number status per gene

per tumor. Both copy number and matching WES data was available for 273 GBM patients.

For copy number data, we remove calls in regions of copy number polymorphism, as called by

the Broad Institute pipelines, and we keep only copy number alterations in genes that are in called

GISTIC2 peaks. For the nucleotide variant calls, we record any gene with a somatic

nonsynonymous mutation as mutated in the patient. The result of this initial step is a binary

matrix of patients and genes that marks patients as having a mutation in a gene.

We combine the two matrices in an "or" gate fashion. Finally, we merge genes on the same

chromosome that are altered in exactly the same samples into a single unit. It is important to note

that copy number aberrations are usually not focal events targeting a single driver gene, and in

fact often involve entire chromosomes. Thus, even distant genes on the same chromosome as

another gene already included in the module will score as the best candidates for module

inclusion, although this does not reflect any functionally interesting genetic interaction (Figure

3-2). In order to remove this bias, we do not allow any module to contain more than one gene

from the same chromosome.

Page 39: Rachel D. Melamed

28

 Figure  3-­‐2  Illustration  of  copy  number  linkages  in  the  GBM  cohort    

The  entire  chromosome  1  is  shown.    Two  significantly  recurrently  amplified  genes  are  shown,  MDM4  and  PRDM2.    While  both  events  are  selected  for  amplification,  and  chromosome  1  is  the  largest  chromosome,  it  is  impossible  to  distinguish  significantly  co-­‐occurring  events  from  the  effect  of  the  amplification  of  the  entire  chromosome.  

3.1.2.2 Scoring the module Our aim is to find the most mutually informative set of genes, using the total correlation score:

𝑇𝐶 𝑋!,𝑋!,⋯ ,𝑋! = 𝐻(𝑋!)!

!!!

− 𝐻 𝑋!,𝑋!,⋯ ,𝑋!

To find the significance of this value, we apply the G-test as follows.

The total correlation, or mutual information, of two variables x1 and x2 can be reorganized to form

the Kullback-Leibler divergence from the joint distribution, p(x1 ,x2) of the independent

distribution, p(x1 )�p(x2). As outlined in (Goebel et al.), we can treat the Kullback-Leilbler

divergence (and therefore mutual information) as a function of the joint p(x1 ,x2), and expand this

as a Taylor series about the point p(x1 ,x2) = p(x1 )�p(x2). The resulting expression, using the

expansion terms only up to order 2, is identical to that of a chi-squared statistic, when multiplied

by N (the number of data points, to convert probabilities to count-equivalents), and 2�ln(2),

accounting for the change of base from base 2, and the coefficient of the Taylor series expansion.

PRDM2& MDM4&

Column1 no(PRDM2 PRDM2no(MDM4 187 10MDM4 47 29

P&=&4x10,12&&

Page 40: Rachel D. Melamed

29

The degrees of freedom of this chi-squared statistic would be 1, in our case of mutual information

of two binary variables. To give an example of the calculation of the degrees of freedom, if we

have two genes in the module, there are four possibilities, which can be viewed as a two by two

contingency table as below:

x2 non-mutated x2 mutated

x1 non-mutated

x1 mutated

This table has 22 = 4 cells with 3 constraints, consisting of the number of mutations per each of

two genes, and the total number of samples. Thus, there is one degree of freedom.

The multi-variable total correlation is an extension of the deduction for mutual information.

Using the same logic as outlined above, we can reformulate the total correlation formula as

where is the Taylor series remainder term of order 3; is the observed number

of events, and is the expected number of events. Again, according

to the chi-square test, approximately follows a chi-squared distribution, with degree

of freedom, for n binary genes mutation statuses, of (Kullback 2012) (correct only

TC(X1,X2,,Xn ) = H Xi( )−H X1,X2,,Xn( )i=1

n

=1ln2

pX1X2Xn x1, x2,, xn( ) lnpX1X2Xn x1, x2,, xn( )

pX1 (x1) ⋅ pX2 (x2 ) ⋅⋅ pXn (xn )xn

∑x2

∑x1

=1

2 ln2

pX1X2Xn x1, x2,, xn( )− pX1 (x1) ⋅ pX2 (x2 ) ⋅⋅ pXn (xn )$% &'2

pX1 (x1) ⋅ pX2 (x2 ) ⋅⋅ pXn (xn )+O3

xn

∑x2

∑x1

=1

2N ln2

n x1, x2,, xn( )− n(x1) ⋅n(x2 ) ⋅⋅n(xn ) / Nn−1$% &'

2

n(x1) ⋅n(x2 ) ⋅⋅n(xn ) / Nn−1 +O3

xn

∑x2

∑x1

O3 n x1, x2,, xn( )

n(x1) ⋅n(x2 ) ⋅⋅n(xn ) / Nn−1

2N ln2 ⋅TC

2n − n−1

Page 41: Rachel D. Melamed

30

when the number of samples is bigger than 2n). As in our example of two binary genes above,

the formula calculates the degrees of freedom as 22 - 2 – 1 = 1.

Actually, total correlation is a special case of the G-test. In statistics, G-tests are formulated as

𝐺 = 2 𝑂! ∙ ln  (𝑂!𝐸!)

!

where is the observed distribution (frequency), and is the expected distribution based on

null assumption. It can be proved that approximately follows a chi-squared distribution (Sokal

and Rohif 1981).

It is important to mention that the number of samples is important to the approximation of the

distribution of total correlation. We simulate five independent variables with different number of

samples ranging from 2 to 100. The theoretical value approaches simulation results very well

when the number of samples is larger than 20, but the G-test fails when sample size is small.

Therefore in our application of our total correlation method, if the number of samples is larger

than the degrees of freedom, , we can use the G-test. Otherwise, we must use a

permutation method to calculate the p-values.

3.1.2.3 Module selection The greedy method starts from the pair of genes with the highest mutual information. To grow the

module from this initial pair of genes, we then test each other remaining gene to find one, which,

together with the existing gene set, will create a set with the highest total correlation. If no

module is found at a greater significance level than .05 divided by the number of genes remaining

in the module, growth is terminated. We continue to add genes until reaching the maximum

feasible module size where joint entropy can be estimated, which is less than the logarithm of the

number of samples.

Oi Ei

G

2n − n−1

Page 42: Rachel D. Melamed

31

The goal of the SA method is to sample modules of genes in proportion to the total correlation of

the modules. The GAMToC SA starts from any initial gene set of a selected size. We use the

maximum feasible module size for G-test calculations, given our sample size. For the GBM

combined copy number and whole exome data set of 273 tumors, this is a module of eight genes.

The chain continues at each iteration by randomly choosing a gene from the module and

replacing it with another gene chosen at random from the non-module genes. If the score of the

module is improved by this replacement, then the replacement is retained. If instead the new gene

creates a decreased total correlation, the module change has a probability of being retained

(paccept), according to the change of the total correlation. We define log (paccept) as proportional to

the change of total correlation, with a proportionality constant that we define as 1/temp.

The temperature starts as "hot", such that a small decrease in total correlation results in a likely

probability of acceptance. The temperature continues to decrease by a percentage after a

minimum number of iterations and a minimum number of changes to the module. After the

change is retained or discarded, the resulting module is the next state in the chain. If the

annealing process stops for a certain number of iterations, it will restart at the highest total

correlation module that was reached in the course of the annealing, and continue the process at

the current temperature. The process will continue until the annealing converges. The final

highest total correlation module is our solution.

Page 43: Rachel D. Melamed

32

 Figure  3-­‐3:  Visualization  of  the  relationship  between  temperature  (in  legend),  change  in  total  correlation  (x-­‐axis),  and  acceptance  probability  in  the  simulated  annealing  (y-­‐axis).  

3.1.2.4 Simulation of module and assessment of results For the simulation, we chose to create a data set of 100 genes and 100 patients, and we embed a

six gene module in this data. Thus, each simulation creates a binary matrix of gene mutations per

patient. For the embedded module, the simulation uses a parameter specifying the fraction C of

the patients that are covered in the module pattern, where the rest of the patients have no module

pattern. The other parameter specifies random noise N added to the module genes.

First, we simulate the background mutations for independently mutated genes. On average in the

glioblastoma data each gene is mutated in 12.9 samples, with a steep decline in number of genes

with higher mutation rates. We model this distribution with an exponential, with the empirical

mean value as the distribution parameter. Sampling from this exponential distribution, we

simulate the background independent mutation rates for each gene. Then we generate the

mutations for each patient for each gene as a Bernoulli process according to that gene's simulated

mutation rate. Next, we embed in this data a module covering C patients. We generate an

exclusive or triplet for the first three genes. We use a multinomial distribution, based on the

mutation frequencies of the three genes, to pick which two of the three would be mutated for each

covered patient. The final three genes are the negation of the first three genes. Then, according to

the noise, N% of the module bits are flipped.

0 0.1 0.2 0.3 0.40

0.2

0.4

0.6

0.8

1

delta−TC

p−ac

cept

0.1 0.05 0.010.0050.001

Page 44: Rachel D. Melamed

33

For each simulation, the greedy module and the simulated annealing module are assessed, and we

compare how many of the 6 genes are recovered in each of the 100 simulations for each

parameter setting (Figure 3-5C).

3.1.2.5 Comparison to tumor classifications After obtaining our results, we compare the genes included in our module to previous

classifications of tumors. This is motivated by the clear correspondence of our module to aspects

of these previous tumor classifications, such as the association of Classical tumors with EGFR

and the G-CIMP with IDH1. Tumor classification performed by the TCGA in (Brennan et al.

2013) was downloaded from http://tcga-

data.nci.nih.gov/docs/publications/gbm_2013/supplement/Molecular_subtype_classification.xlsx.

Of the patients included in our study, 233 were classified in that work. We compared these

classifications with mutation status of each module gene, in order to assess whether the mutations

were markers of GBM subtypes.

3.1.3 Results 3.1.3.1 Utility of searching for mutually informative gene sets While many well-characterized cancer driver genes are highly recurrent, more rarely mutated

tumor drivers are difficult to identify amidst unstable genomes when using mutational frequency

alone. Thus, we must utilize other aspects of the alteration pattern of these genes, such as mutual

exclusivity or co-mutation with other genes, keeping in mind that frequency of individual lesions

may be low.

As can be seen in Figure 3-4A the number of samples needed to statistically identify mutual

exclusivity between a pair of genes grows large when the frequency of mutation is low, and this

size is orders of magnitude larger than the number needed to identify co-mutated pairs. This is

intuitive as the expectation is that two infrequent mutations are most likely to have no co-

Page 45: Rachel D. Melamed

34

occurrence. When a set of mutually exclusive genes, each with the same low mutational

frequency, is instead assessed for a significantly related mutation pattern, the number of samples

required to attain significance is much lower (Figure 3-4B).

Figure  3-­‐4:    Ability to find multi-gene co-mutational patterns.

A

Mutually exclusive mutation set

B

C

samples Total correlation p-value = .004 Pairwise correlation p-value = .76

Mutually exclusive mutations

Co-occurring mutations

gene

s

Page 46: Rachel D. Melamed

35

A. Finding mutually exclusive pairs of gene mutations requires orders of magnitude more samples as compared with finding co-mutated genes. B. With a larger set of genes, fewer samples are needed. C. For an exclusive or triplet pattern, the total correlation is strong, but a pairwise correlation or anti-correlation score would fail to detect a relationship.

Additionally, multi-gene patterns may exist other than mutual exclusivity or co-occurrence. An

example would be an “exclusive or” triplet of genes where lesion of any two of the genes is

enough to change a phenotype, and the third adds no further advantage. As is shown in Figure

3-4C, the total correlation of this three-gene pattern is highly significant, but the genes display no

mutual exclusivity or co-occurrence pattern.

3.1.3.2 Evaluation of greedy and simulated annealing algorithms We have implemented two methods that integrate copy number and point mutation data to find

sets of genes with high total correlation, both taking different approaches to finding patterns in

this data. The greedy method finds a module by starting from the pair of genes with the strongest

mutual information, iteratively adding the gene that creates the best score. On the other hand, the

SA method allows us to explore the broader landscape of modules in order to find an optimal

solution. In general, SA methods semi-randomly sample possible solutions to a hard problem,

sampling those with the better scores (objective function) more often. Our application of SA

samples combinations of genes with high total correlation, and it can find a solution with a higher

score. A detailed description can be found in the Methods section.

First, we compared the running time of our implementations against each other and against an

exhaustive method. We create a simulated data set containing 100 genes and 100 samples. As

shown in Figure 3-5A, time complexity of the exhaustive method increases exponentially with

module size, while the greedy method will finish in tens of seconds and the SA method will finish

in tens of minutes.

Page 47: Rachel D. Melamed

36

To evaluate the accuracy of the greedy and SA approximations, we randomly generate an

embedded module in randomly simulated data, as described in 3.1.2.4. This simulated module has

a 6 gene pattern including an exclusive or triplet of genes and their negations (Figure 3-5B),

while all other genes are randomly mutated at an exponentially distributed background mutation

rate. Two simulation parameters are used: coverage and noise. In a larger coverage, most patients

contain this pattern for the module genes, while the rest of the

Figure  3-­‐5:  Comparison of different methods in GBM mutation data.

A. Time complexity of SA, Greedy method and Exhaustive method, as compared to the increase of module size. B. Example of a simulated module (with coverage 50% and noise 5%). C. The average number of simulated

A

C

1% 5% 15%0

1

2

3

4

5

6

noise

Mea

n no

. mod

ule

gene

s re

cove

red

(of 6

)

20% covg,greedy20% covg,SA50% covg,greedy50% covg,SA80% covg,greedy80% covg,SA

3 4 5 6

10

100

1000

10000

100000

1000000

seco

nds

module size

B

g6 g5 g4 g3 g2 g1

3 4 5 6

10

100

1000

10000

100000

1000000

module size

seco

nds

Simulated AnnealingGreedyExhaustive

Page 48: Rachel D. Melamed

37

module genes recovered (out of the full 6 gene module) across 100 simulations. The SA method has better recovery than the greedy method, but both recover 5 of the 6 genes on average at 50% or more coverage.  

patients have a pattern as generated by the background model. Thus, the score of the module

genes will be higher and the module will be more readily detected. At each coverage, the noise

varies from low noise (on average 1% of the mutation statuses are flipped at random), to high

noise (15% of the mutation statuses). We generate the module and the rest of the data 100 times

for each setting of the parameters. Then, we assess the average number of genes from the gene set

that is recovered by the algorithms, where 6 genes is the maximum (Figure 3-5C). Note that in

each setting, including low coverage and high noise, at least three of the six module genes are

recovered.

3.1.3.3 Application of Greedy GAMToC to TCGA GBM samples First, we explore modules of different sizes using only the mutation data, which is much more

sparse than copy number data. The resulting mutation matrix contains 256 genes that are mutated

in at least 2% of 283 patients with whole exome sequencing. For a module of size three, the

simulated annealing method and the greedy method arrive at the same module of mutated genes.

Comparing this against the exhaustive method, we find that GAMToC recovers the best module

in the data. When module size equals four, it would take 3.5 days for the exhaustive method to

search all modules (Figure 3-5B).

Notice that while total correlation increases according to the module size, it does not make sense

to compare different size modules in terms of total correlation. We use the G-statistics to

overcome this issue (refer to method for detail), and calculate p-values based on the chi-square

distribution for all modules. We find that the five-gene module containing TP53, IDH1, ATRX,

RB1, and PTEN is the most large and significant one in this example (Figure 3-6). In fact, TP53,

IDH1, ATRX, RB1 are all significantly positively correlated with each other. PTEN has a

significant negative correlation with IDH1, as well as a positive correlation with mutation in RB1.

Page 49: Rachel D. Melamed

38

 Figure  3-­‐6:  Recovery of different module sizes in only mutation data.

The p-value associated with the total correlation is indicated on the y-axis, and the modules for each size are shown. For each size, the same module was found from the greedy and SA methods. Blue edges represent negative correlations between the genes, while red edges are positively correlated. Edge thickness denotes the strength of the association. Node size represents the frequency of alteration. Node border width represents the number of nonsynonymous mutations in that gene.  

Next, we apply the greedy algorithm to a set of 273 tumors from the TCGA GBM project that

have available copy number and exome sequence. Collating these data results in a mutation

matrix of 756 alterations on the 273 samples. The greedy module recovered displays an

interesting pattern of pairwise co-occurrence and mutual exclusivity between mutations (Figure

3-7A). It is important to note that total correlation finds a multi-gene structure of related

alterations: as in the "exclusive or" example (Figure 3-4C), there may not be any strong pairwise

relationships in a strong module. However, for visualization purposes we display the resulting

modules in terms of their network of pairwise positive correlations (co-occurrence of a pair of

genes) and negative correlations (mutually exclusive mutations). Thus, for the remainder of this

work we provide a pair-based network visualization of the module structure.

3 4 5 6

1.00e−05

1.00e−10

1.00e−15

module size

TC p−v

alue

RB1

IDH1

TP53PIK3R1

PTENATRX

PTEN

IDH1

TP53

ATRX

RB1

IDH1

TP53

ATRX

RB1

IDH1

ATRX

TP53

Page 50: Rachel D. Melamed

39

Figure  3-­‐7:  Networks of total correlation modules.

The legend is the same as in Figure  3-­‐6, except for that node color represents average copy number amplification (red) or deletion (blue). A. The greedy module from glioblastoma. B. The Seed-GAMToC module, seeded with CDK4. C. The SA module. D. The genes from the greedy and SA modules are compared to subtypes of glioblastoma. The darker the red, the stronger the association (Fisher's exact test) of gene mutation status and that subtype.

We grow a greedy module up to the maximum feasible size, which is eight genes. In the greedy

module, patients appear more likely to display mutations that co-occur with TP53, IDH1, and

RB1, or that are mutually exclusive with these genes. Patients with mutation or deletion of TP53

are significantly more likely to also have mutations in IDH1 and ATRX, and ATRX and IDH1 as a

pair have the highest mutual information in the data set. The deleted and mutated gene RB1

strongly co-occurs with TP53 lesions, though it has no positive correlation with IDH1 or ATRX.

Deletion to the terminal section of chromosome 11p, which GISTIC2(Mermel et al. 2011)

identifies as peak gene BRSK2, also frequently co-occurs with lesions of TP53 and RB1. The

11p15 region is imprinted, and it is known to be deleted, to undergo loss of heterozygosity, and to

BRSK2CDKN2A (region)

RB1

TMCO5A

NKAIN2

CD33

TP53

PTPN21

IDH1

CDKN2A (region)

ATRX ADARB2

EGFRBRSK2

RB1

TP53 RB1

TP53

TMCO5A

BRSK2

CDKN2A (region)

SPTA1

CDK4 (region)

PAPLN

A

D

B

C

-log1

0 p-

valu

e

TCOM5A CDKN2A (region) EGFR ADARB2 KNAIN2 CD33 PTPN21 RB1 BRSK2 TP53 ATRX IDH1

G-C

IMP

Pro

neur

al

Mes

ench

ymal

Neu

ral

Cla

ssic

al

Page 51: Rachel D. Melamed

40

have differential epigenetic regulation in multiple cancer types (Schwienbacher et al. 2000;

Onyango and Feinberg 2011).

Many of the genes that co-occur with TP53 alteration have a mutually exclusive pairwise

relationship with copy number alterations in EGFR, CDKN2A region, or chromosome 10

deletion. The dominant effect of chromosome 10 deletion is likely the inactivation of the tumor

suppressor PTEN, which is one of the most prevalent events across tumors. However, it is

interesting that a large section of the chromosome is deleted, and not just PTEN. The greedy

GAMToC selects the GISTIC2 deletion peak on the terminus of 10p, containing ADARB2, as

well as IDI1, IDI2, and WDR37. Very importantly, this region has stronger pattern of positive

correlation with EGFR deletion, and negative correlation with IDH1 mutation, than does PTEN

deletion, explaining its selection by the greedy method. While the full module of eight genes is

very interesting, the seven gene module (removing CDKN2A region) is more statistically

significant.

3.1.3.4 Seeding the greedy algorithm The greedy method has a disadvantage of performing only a local search for a high scoring

module. It starts from the pair of genes with highest mutual information (pairwise total

correlation), and uses a greedy approach to find a module that contains that pair. While we also

develop the SA method to find other modules, the greedy method has two advantages for

understanding cancer evolution. First, exploring the search space around the pair of genes with

the highest mutual information is informative of processes in cancer, as we show above. Second,

the greedy algorithm allows us to choose the starting point of the module search, by fixing an

initial gene, that we call a seed gene. In this procedure, termed Seed-GAMToC, we identify a

local maximum of total correlation that includes that seed gene. First, we find the partner gene for

the seed gene. forming a gene pair with the highest mutual information, and we grow the greedy

Page 52: Rachel D. Melamed

41

module from this pair. Thus, we seek to characterize a given gene by finding what module of

high total correlation contains that gene, or, in other words, the genetic context in which

mutations of that gene appear. Discovering these relationships, such as the genetic context in

which disruption of a query gene is advantageous, can illuminate the function of putative cancer

genes.

Among the results of cancer genomics studies are frequent mutations in genes with a role in the

cancer of interest that is not fully characterized. We run Seed-GAMToC for a number of genes

that are significantly mutated or in copy number peaks in GBM patients, but were not selected by

the greedy algorithm. We were interested in CDK4 because it is a cell cycle kinase that is focally

amplified in GBM, and mutual exclusivity has been observed between amplification of CDK4,

deletion to the CDKN2A locus, and deletions and mutations to RB1. We wondered what factors

influence this mutual exclusivity, and we ran Seed-GAMToC starting from CDK4 (Figure 3-7B).

In fact, while CDKN2A is mutually exclusive with both CDK4 and RB1, the latter as a pair are not

strongly mutually exclusive (chi-square p-value = 0.39). However, in patients with no CDKN2A

deletions, their conditional mutual exclusivity is significant (chi-square p-value = 4x10-4). It is

interesting that both CDK4 and RB1 have strong co-occurrence with other genes that are also

mutually exclusive with CDKN2A. CDK4 co-occurs in patients with mutation to SPTA1, a

recurrently mutated member of the spectrin cell scaffolding complex. Mutation to SPTA1 could

impact cell adhesion, and mutations to other spectrins have been shown to affect cell cycle

regulation(Metral et al. 2009). On the other hand, RB1 co-occurs with TP53 and its correlated

genes. CDKN2A can regulate CDK4 and RB1, as well as TP53, explaining this discovery.

Because RB1, CDK4, CDKN2A all have roles in cell cycle, we also looked at the patterns

associated with other significantly mutated cell cycle genes. For example, CDK6 plays a similar

Page 53: Rachel D. Melamed

42

role in promoting cell cycle progression as CDK4, and, like CDK4, this gene is strongly

amplified. Seeding with CDK6, we find a strong correlation with PTEN deletion, and anti-

correlation with ATRX and IDH1 mutation (Figure 3-8A). Thus, unlike CDK4, CDK6 may be a

beneficial amplification in the context of the mitogenic PI3-kinase pathway, which is deregulated

by PTEN deletion or mutation. On the other hand, another mitogenic event, amplification of

PIK3C2B (along with its chromosomal neighbor MDM2), seems to cooperate with deletion of

RB1 and amplification of the cell cycle promoting amplification MYCN (Figure 3-8C). One final

gene closely related to cell cycle regulation is CCNE1, and amplification of this gene is strongly

mutually exclusive with TP53 alteration(Figure 3-8B). One effect of TP53 inactivation is in fact

de-repression of CCNE1, and CCNE1 likewise can mediate genetic instability(Hwang and

Clurman 2005). Thus, the module identified by the greedy method is useful for understanding the

role of a query gene in glioblastoma development, including closely functionally related genes.

 Figure  3-­‐8:  Networks seeded with query genes.

A. CDK6 seed. B. CCNE1 seed. C. PIK3C2B seed (co-amplified with MDM2).  

PTEN

RB1

TMCO5A

TP53

BRSK2

ATRX

IDH1

CDK6

PAPLN

CCNE1

TMCO5A

NKAIN2

TP53

BRSK2RB1

LSAMP

A B

RB1

GABRA4

BRSK2

PIK3C2B

KIF16B

MYCN

TP53

C

Page 54: Rachel D. Melamed

43

3.1.3.5 Simulated Annealing results consistently identify a high-scoring module The SA algorithm provides an alternate mode of selecting a module, allowing us to more broadly

search for a high scoring module. Unlike the greedy method, SA can escape local maxima and

find a higher scoring module. Over the course of the semi-random sampling, the SA undergoes

“annealing”, becoming more selective for high total correlation modules. A run of SA will

eventually converge on one module, but in practical settings SA will converge on a local

optimum. Because there are many more copy number events than nucleotide mutation events, and

all alterations are counted equally in GAMToC, the SA is more likely to converge on states

involving broader copy number changes, making it somewhat less sensitive to mutational patterns

or very focal SCNAs than the greedy algorithm. In multiple runs of the SA, one best module was

found, that has a higher total correlation score than the greedy module (1.28 as opposed to 1.03),

and is extremely statistically significant.

In the SA’s best module, a pattern appears that is related to that of the greedy module, but

dominated by copy number changes (Figure 3-7C). As with the greedy module, the SA module

has a set of genes that co-occur with mutation of TP53. This includes, as before, RB1 and BRSK2.

Additionally, deletion in chromosome 15, in GISTIC2 peak gene TMCO5A co-occurs with these

genes, while another deletion region on chromosome 14 centered on PTPN21 is also associated

with some of TP53's co-occuring partners. Mutually exclusive with TP53 and RB1 mutations is

again deletion to the CDKN2A/CDKN2B locus.

3.1.4 Discussion Our algorithms search for genes with related occurrence of alteration across tumor samples, based

on the premise that the joint alteration status of genes in tumor samples can inform us of the

evolutionary process behind the cancer. Unlike mutual exclusivity methods that impose a single

structure on the data, our approach is able to form a more comprehensive picture of alteration

patterns that exist in cancer data. The result of applying GAMToC to the TCGA GBM data is a

Page 55: Rachel D. Melamed

44

network of genes with a jointly related mutation pattern, suggesting that the alterations in GBM

do in fact follow an underlying structure. The interpretation of the module can be more complex,

as opposed to mutual exclusivity, which is often interpreted as representing alternative mutations

in a pathway. But one interpretation is that the co-occurring sets of gene lesions represent

alternative pathways to glioblastoma development: there are different contexts in which these

different lesions provide a selective advantage.

The interpretation of the sub-module structure as indicating routes to GBM development suggests

that patients harboring different sets of mutations may have different characteristics. In fact, this

pattern has been observed in the TCGA GBM cohort. Subtypes of glioblastoma have been

identified by expression (Verhaak et al. 2010), as well as by methylation(Noushmehr et al. 2010),

and these have been related to specific genetic alterations(Brennan et al. 2013). Patients with a

methylation profile known as Glioblastoma CpG Island Methylator Phenotype (G-CIMP) have

better survival, while patients with a gene expression pattern that follows the Proneural subgroup

have different response to therapy. To support the hypothesis that the GAMToC module is

indicative of these types of tumors, we examine if the GAMToC modules are related to these

patient subtypes. We test whether patients with mutations to each module gene are more likely to

fall into one of the subtypes. In result, the Classical and Proneural gene expression subtypes are

strongly associated with certain module genes, as is the G-CIMP methylation group (Figure

3-7D). Thus, our approach successfully captures biological differences between patient groups, as

reflected in different patterns of genetic lesions.

The Classical subtype typically has co-occurring mutations in EGFR and CDKN2A. Mouse

models have suggested that activation of EGFR can cooperate with loss of the CDKN2A locus

and PTEN to generate gliomas with high resemblance to GBM (Zhu et al. 2009). However,

Page 56: Rachel D. Melamed

45

rather than PTEN, the chromosome 10 deletions of ADARB2 are selected by GAMToC. This

region is strongly co-deleted with PTEN (chi-squared p-value = 3.7x10-25), since in many cases of

PTEN deletion most of chromosome 10 is deleted. However, ADARB2, IDI1, IDI2, and WD47

have a stronger pairwise pattern with the other module genes chosen by GAMToC. Additionally,

patients with this deletion are significantly more likely to fall into the Classical expression

subtype (chi-squared p-value = .029), while PTEN is weakly associated with the Mesenchymal

subtype (p-value = .086). Thus EGFR amplification, chromosome 9 deletion of CDKN2A and

CDKN2B, and ADARB2 locus deletion (including IDI1, IDI2, and WDR37) are all negatively

correlated with TP53 and are all associated with the Classical expression profile.

In contrast to the better understood Classical subtype of GBM, the IDH1-p53 network associated

with G-CIMP and with Proneural groups has been long studied but has so far remained of

uncertain significance for tumor initiation in the brain. The strong co-occurrence of TP53

alterations with deletions of 11p15 (BRSK2) and 15q14 (TMCO5A) is an exciting novel finding.

While TP53, IDH1, ATRX, and BRSK2 are all highly associated with G-CIMP, TP53 and BRSK2

are also strongly associated with Proneural status. BRSK2 is particularly intriguing because it is a

kinase that is highly expressed in brain and may be involved in apoptotic stress response(Y.

Wang et al. 2012) and cell cycle regulation(R. Li et al. 2012). Proneural tumors are also strongly

associated with TMCO5A deletion, a lesion that, distinctively, is not associated with G-CIMP

tumors. The genes in these regions may provide the missing element to recapitulate the

gliomagenic process in these tumors.

It is also interesting to compare our modules with modules of mutually exclusive genes. Methods

to find patterns of mutual exclusivity, such as MeMo (Ciriello et al. 2011) or DENDRIX(Mark D.

M. Leiserson et al. 2013), have pointed out genes also selected by GAMToC. These methods

Page 57: Rachel D. Melamed

46

sometimes claim to find new pathway interactions in this manner, exemplified by the mutual

exclusivity between CDKN2A, CDK4, and RB1, or between CDKN2A and TP53. But GAMToC’s

ability to find other relationships between mutations shows that the mutual exclusivity is related

to the subtype-specific nature of mutations. It is very interesting to focus on the example of the

retinoblastoma pathway, which can integrate signals from the mitogenic pathways (PI3-kinases,

PTEN), and DNA damage (TP53), among others. We find that mutations to the DNA damage

(TP53), cell cycle (RB1), and mitogenic (PTEN) pathways are prevalent across the glioblastomas,

but that different specific alterations seem to confer subtle advantages in different mutational

backgrounds. In Figure 3-9 we outline the subtype associations of genetic alterations affecting

these pathways. For example, TP53 and RB1, as well as CDK4, are advantageous for G-CIMP

and proneural tumors, while CDKN2A is a dominant lesion in classical glioblastomas, and

CCNE1, and CDK6 also occur less frequently in the Proneural tumors. Highly functionally

related genetic alterations have been suggested to have similar effects. In the case of CDKN2A

(p16) and TP53, both lesions alter DNA damage response, while cell cycle regulation is

transformed by mutations to CDKN2A, CDK4, CDK6, CCNE1 and RB1. However, far from the

simplifying assumption that mutually exclusive events represent alternative equivalent routes to

cancer development, clearly there are subtleties resulting in subtype-specific mutations. The data

imply that mutations to genes in the same pathway are not in fact interchangeable.

Page 58: Rachel D. Melamed

47

Figure  3-­‐9  Cell  cycle,  DNA  damage,  and  mitogenic  gene  subtype  associations.  

A. Fisher's exact test is used to assess association of mutation status with each subtypes. Darker red is stronger association. B. In reverse, this shows the depletion of mutation status of the gene in each subtype. C. A schematic of the functional relationships between genes involved in RB1 regulation. The fill of the boxes represents prevalent amplification or deletion of that gene in glioblastoma. The line on the outside of the boxes represents the subtype specificity of the gene, as calculated for part A.  

!log10'p!value'

CDKN2A'

RB1'

MYCN'

PTEN'TP53'

!log10'p!value'

CDK4'CDK6'

CCNE1'

PIK3C2B'(MDM2)'

!log10'p!value'

p14' p16'Classical'Proneural/G!CIMP'

Proneural'

A'

B'

C'

Page 59: Rachel D. Melamed

48

More generally, our results also provide insight into the nature of subtype-specific lesions. As the

method will detect any non-random pattern of alteration in a collection of samples, the resulting

module may contain genes that are co-mutated because they are both present in tumors of the

same subtype or environmental condition, rather than because of any direct functional interaction.

While patterns of joint lesion status do not allow us to distinguish between these two conditions,

our results show that genetic context has a strong influence on selection. Thus, the distinction

between subtype-specific co-alteration versus synergistic co-alteration may be thought of as a

matter of the degree of selective advantage, rather than as two different phenomena. In

conclusion, we have developed a method to uncover novel relationships between genes that are

key to cancer development, and we have related the findings to previous subtypes of

glioblastoma. Understanding the combination of genetic alterations present in patients with a

tumor will help to target therapies to their pattern of aberrations. This application is an example

of the power of a generalized entropy-based approach to gene set recovery.

3.2 GAMToC-L: Using patterns of co-selection of cancer genes to

identify and contextualize novel drivers

The results from GAMToC (3.1) were highly encouraging: genes such as BRSK2, that have not

been highlighted in work in glioblastoma, stood out in this analysis. This gene is recurrently

deleted—GAMToC relies on recurrently altered genes for input to the analysis. However, many

genes are recurrently altered in glioblastoma, but very few have such a strong pattern of joint co-

occurrence and mutual exclusivity. Therefore, the method is able to highlight potential driver

genes with an evident pattern of selection that is not reliant only on recurrence.

Page 60: Rachel D. Melamed

49

We wondered if this idea could be extended: perhaps rather than starting from a list of recurrently

mutated genes and finding those among them that have a non-random relationship, we could start

from all copy number and point mutations, and use the total correlation to both identify driver

genes and to find a module pattern. Thus far, most studies examining relationships between genes

have started from the recurrent mutations because these represent the most likely driver genes

(Ciriello et al. 2011; Ciriello et al. 2013; Vandin, Upfal, and Raphael 2012; Akavia et al. 2010).

Because every cancer is different, it is well known that some low-frequency drivers will never be

captured by recurrence-based measures, even with large sample sizes(Lawrence et al. 2014; Mark

D M Leiserson et al. 2014; Torkamani and Schork 2009). Particularly, this has left copy number

data as an under-utilized resource in cancer genomics. There are so many copy number alterations

in a given tumor that they clearly cannot all be important events for the tumor. But copy number

changes are major events with a demonstrated high impact on gene expression and cellular

function. Novel methods to find important changes, both in copy number and in nucleotide

sequence changes, are strongly needed.

Total correlation could provide a new signal for positive selection in cancer. A module of genes

cannot have a strongly non-random pattern when the component genes are extremely rarely

mutated. However, presence of a strong module pattern that includes a gene could allow us to

distinguish the likely drivers among genes that are altered at similar frequencies. In this section of

my dissertation, I describe a new method, called GAMToC-L, (GAMToC-Landscape), that is

able to explore the space of high total correlation modules and identify more subtle patterns of

selection in a number of cancers.

This work is based off of the simulated annealing GAMToC method, with a few important

changes. First, GAMToC only found modules among recurrent genes, limiting its search to 256

Page 61: Rachel D. Melamed

50

genes in the case of the TCGA glioblastoma data. GAMToC-L instead uses all genes with any

copy number or point mutation, a set that comprises from 6000 to 12000 genes per tumor type.

Second, while GAMToC only considered one high scoring module, discovered through two

different search methods, GAMToC-L examines the distribution of modules generated by the

simulated annealing method.

 I extend the results from GAMToC by applying the new method not only to glioblastoma, but

also to lower grade glioma. This provides an interesting new contrast to the glioblastoma results,

as lower grade gliomas share cells of origin with the higher grade glioblastoma, and the higher

grade tumors sometimes arise from lower grade. I show that novel, but highly plausible, driver

genes are uncovered, and patterns of co-mutated genes provide further insight into the biology of

these cancers. This work is in preparation for submission.

 3.2.1 Methods 3.2.1.1 Relationship with GAMToC As mentioned, GAMToC-L is heavily based off of the methods of GAMToC, as described in

3.1.2.3. Briefly, in GAMToC, a binary input matrix of samples by mutated genes is created. The

first major difference between GAMToC and GAMToC-L is that in GAMToC-L the input is not

restricted to recurrently mutated genes, but this matrix contains all genes that are mutated in more

than f patients. As mentioned in the introduction, this increases the search set from 256 genes to

6120 genes mutated in three or more patients in the GBM data.

GAMToC’s simulated annealing procedure starts from a randomly generated module. It is

important to note that we apply the same simplifying rule (see Figure 3-2) that allows only one

gene per chromosome in the module. The module size is determined by the number of samples

available: with N samples, only log2(N) = M binary variables can possibly be observed in all of

their states, so this represents the absolute upper limit on module size. In order to capture

Page 62: Rachel D. Melamed

51

complex relationships, we use this size for M. At each step of the simulated annealing, we

randomly replace a gene from the current module, choosing a random replacement gene. This

change is retained if it creates a higher total correlation, and otherwise it is probabilistically

retained, and the probability is tuned by a temperature parameter t. Tests show that the initial

temperature does not affect results, provided it is high enough (greater than .1). After a number of

iterations, i, and a number of changes to the module, c, the temperature is lowered by a

percentage, p. Over the course of the iterations, the temperature decreases and the total

correlation increases to a plateau (Figure 3-10).

Figure  3-­‐10  Effect  of  decreasing  temperature    

A.  Over  the  iterations  of  the  simulated  annealing  (x-­‐axis),  the  temperature  (blue  line)  decreases,  and  the  total  correlation  (green  line)  increases,  and  its  variance  decreases.    This  example  comes  from  the  TCGA  GBM  cohort.  The  dashed  line  shows  the  total  correlation  attained  by  the  greedy  method,  applied  to  the  same  data.      B.  Reproduced  from  3.1.2.3,  as  temperature  decreases  the  probability  of  accepting  a  change  that  lowers  the  score  also  decreases.  

The simulated annealing will reach a local maximum, at a low temperature, defined by no change

in u iterations. Then, if the maximum total correlation in the search space that was explored is not

also the local maximum, the search will restart at the maximum previous value. The result of

these iterations is a distribution in the space of modules. In GAMToC-L we use as much of this

low temperature and high total correlation search space as is feasible. Thus, the distribution of

modules over the module search space, a sort of metadata on patterns of mutation across tumor

cohorts, becomes the input data for GAMToC-L. We call this metadata the module data.

Tempe

rature)

0 2 4 6 8 10 12 14x 105

0

0.05

0.1

0.15

0.2

0 2 4 6 8 10 12 14x 105

0

0.5

1

1.5

2

Total)correla-on)

Itera-on)

Greedy)TC)

0 0.1 0.2 0.3 0.40

0.2

0.4

0.6

0.8

1

delta−TC

p−ac

cept

0.1 0.05 0.010.0050.001

A) B)

Page 63: Rachel D. Melamed

52

3.2.1.2 Low-temperature module space appears to represent fluctuations around one solution

The results shown in Figure 3-10A are illuminating for a number of reasons. As the temperature

gets very low, the module space explored seems to narrow around a resulting total correlation

score. This space seems to be distinct from that reached in the greedy search method: the total

correlation remains higher than the greedy module total correlation throughout the final iterations.

It is intuitive that if a true module with at least M genes exists in the data, the module space

explored will narrow to this true module: each step only changes one of the M genes, so any

change that improves the total correlation score will improve it by selecting a gene with a

relationship with the existing module genes. At the low temperature, any other gene will rarely

be selected. Thus, GAMToC’s simulated annealing procedure will select a space of modules that

are related to each other.

0 2 4 6 80

100

200

300

400gene frequencies

frequencies (log10))0.5 1 1.5 2 2.5 3 3.50

100

200

300

400

500gene num partners

number of partners (log10))

0 1 2 3 4 5 60

0.5

1

1.5

2 x 104 pair frequencies

number of modules in which pair appears (log10))

A

C

B

Page 64: Rachel D. Melamed

53

Figure  3-­‐11  Distribution  of  the  frequency  of  genes  and  gene  pairs  appearing  in  the  module  data.    

A.  The  log  distribution  of  frequency  of  genes  has  an  approximate  normal  distribution  (kstest  p  =  .97),  indicating  a  highly  skewed  distribution  of  genes  selected  in  the  module  data.  B.  The  number  of  partners  per  gene,  which  is  the  number  of  other  genes  that  appear  in  any  module  with  a  given  gene,  seems  to  follow  an  exponential  distribution.  C.    The  log  distribution  of  the  number  of  times  each  pair  of  genes  appears.    

Interestingly, less than one quarter of the genes included in the input data are present in any of the

one million modules resulting from the final one million iterations of the simulated annealing,

and these genes are chosen in a highly skewed distribution (distribution shown in Figure 3-11A).

Examination of how frequently pairs of genes are chosen together in the same module provides

further support that GAMToC-L converges to a set of related genomic alterations. For each pair

of genes we examine how many modules contain both genes of the pair (distribution in Figure

3-11C). The more frequently a pair of genes appears in the same module, the stronger is the total

correlation of modules involving these genes. If more than one module exists in the data, we

would expect to see a cluster of gene pairs that are frequently co-selected together. We create a

visualization of the gene pairs in the GBM data in Figure 3-12. The visualization indicates that

generally, some genes are chosen more in modules with many other genes. With a few interesting

exceptions, frequently chosen genes co-occur with a broad set of other genes in the module data.

Page 65: Rachel D. Melamed

54

Figure  3-­‐12  Frequency  of  co-­‐selection  of  pairs  of  genes  in  the  module  data.    

Each  row  and  column  is  one  gene,  arranged  in  chromosomal  coordinate  order.  Chromosome  numbers  are  labeled  on  the  x  and  y  axes,  and  chromosomes  are  ordered  for  visual  clarity.  Columns  containing  genes  on  the  same  chromosome  are  outlined  in  the  colored  rectangles.  Only  chromosomes  and  genes  selected  in  the  low-­‐temperature  module  data  are  shown.  Each  point  is  a  pair  of  genes,  with  the  darkness  of  the  point  showing  how  frequently  the  pair  is  co-­‐selected  in  the  module  data.  For  example,  TP53  (arrow)  is  the  only  gene  on  chromosome  17  chosen,  and  it  is  chosen  frequently  with  almost  every  other  gene  in  the  module  data.  This  can  be  viewed  on  the  figure  as  the  dark  vertical  and  horizontal  stripe  of  densely  packed  points  indicating  all  genes  that  TP53  is  chosen  with  (zoom  in  for  best  view).    Note  the  lack  of  pairs  of  genes  chosen  together  within  a  chromosome,  which  is  by  design.  

 

 

1 17 6 5 14 22 15 20 11 9 13 19 4 12

1

17

6

5

14

22

15

20

11

9

13

19

4

12

TP53

Page 66: Rachel D. Melamed

55

3.2.1.3 Identification of consistently and recurrently selected genes We develop a method termed window-validation to identify the recurrently selected genes

consistently chosen across iterations. Like cross-validation, the window-validation compares

results in different subsets of a data set. In this case, the data is the module data resulting from

the simulated annealing. Across the low-temperature iterations, the module space explored will

vary: modules in nearby iterations will be more similar to each other than modules in more distant

iterations. Thus, we use a sliding window approach to split up the data. Each window is a subset

of consecutive iterations, and the makeup of the modules in the window is compared to the

makeup found in the rest of the data.

For a given subset of the module data, for each chromosome, we identify the genes on the

chromosome that are chosen at an elevated rate as follows. Let t represent the number of times a

module in the subset includes a gene from the chromosome, and x represent the total number

genes from that chromosome selected in the subset. Each gene on a chromosome is chosen

mutually exclusively, by design, so the expected distribution of number of times each gene will

be selected, given that a module contains a gene from the chromosome, is multinomial, with a

uniform probability of 1/x for each gene. Thus for an individual gene, its probability of being

chosen versus not being chosen would then be binomial with the same probability. We identify a

the 95-percentile of the binomial distribution with number of trials t, and probability 1/x, in order

to identify a cutoff for the number of times a gene is chosen that is more than expected.

We apply this procedure to the subset of the data in the window, and to the subset of the data not

in the window, and we test whether there is a significant overlap in the genes chosen. If so, this

indicates that the same genes are consistently and recurrently identified in the simulated

annealing. This procedure is applied across all sliding windows. The genes (or localized

Page 67: Rachel D. Melamed

56

chromosomal regions, generally) that are consistently and recurrently selected across the sliding

windows form our resulting module.

3.2.1.4 Iterative module selection We observe that a number of our greedy module genes from GAMToC are not selected in

GAMToC-L. Indeed chromosome 7 (EGFR) and chromosome 10 (PTEN, ADARB2) are never

chosen in the low-temperature iterations. We have applied an iterative approach to finding

multiple modules of interest in the data. All genes that were chosen in the low-temperature

iterations are removed from the data. Additionally, any copy number alterations that are on the

same chromosome as these genes are removed. This will prevent the same relationships from

being re-discovered. Next, the procedure is re-run with the remaining genes. The subsequently

selected module genes may have included strong interactions with the genes that were removed,

limiting the relationships that can be discovered in this fashion. However, this procedure allows

us to find more patterns in the data.

3.2.2 Results 3.2.2.1 Results in glioblastoma data The glioblastoma data provide an interesting result and point of comparison between GAMToC

and GAMToC-L. As mentioned in the Method, we find that in the top module of GAMToC-L,

the total correlation is substantially higher than the GAMToC greedy module. In fact, half of the

genes from the greedy module are not present in GAMToC-L’s top module. In partial agreement

with the GAMToC greedy results, we find two subsets of genes with a mutually exclusive

relationship (Figure 3-13). One set of genes is associated with TP53 and RB1, while the other set

is anticorrelated with these genes. Results include some genes identified via recurrent copy

number alterations or point mutations, as well as some genes that are not recurrent enough to be

identified on their own.

Page 68: Rachel D. Melamed

57

Figure  3-­‐13  GAMToC-­‐L  module  for  the  GBM  data.    

Legend:  node  size  represents  how  frequently  the  locus  is  chosen  in  the  module  space.    Edge  transparency  also  represents  how  frequently  a  pair  is  chosen.    The  width  of  the  edge  represents  the  strength  of  mutual  information  between  a  pair.    Red  edges  are  positively  correlated,  while  blue  edges  are  anticorrelated  pairs.  The  color  of  a  node  represents  its  level  of  amplification  or  deletion  in  the  cohort.    The  node  border  (here,  only  visible  on  TP53)  represents  the  number  of  point  mutations  in  the  node.  

In the TP53-associated group are a diverse set of co-occurring deletions. One of the most

frequently chosen is the deletion in chromosome 11. This locus contains the recurrent deletion of

@6q15

NAA15 FBXW7 @4

GPR132 RCOR1 BCL11B CDCA4

@14

OSBPL8 @12

LCTL MAP2K5 PIAS1 @15

BRSK2 @11

@19pDIS3 RB1 @13

TP53 @17

CDKN2A @9 @20q

@1q25

Page 69: Rachel D. Melamed

58

BRSK2: GAMToC identified this as a TP53-associated event, and the same region is also selected

in GAMToC-L. Interestingly, GAMToC-L’s most recurrent selection in chromosome 11 is

BRSK2, indicating that the deletion identified by GISTIC2 is also the deletion with the strongest

total correlation pattern. As mentioned in 3.1.4, the brain specific kinase BRSK2 is expressed in

brain and may be involved in apoptosis as well as cell cycle regulation (Y. Wang et al. 2012; R.

Li et al. 2012). The module pattern suggests that one of the main effects of alteration in this

kinase in glioblastoma may be its collaboration with TP53 and RB1. Chromosome 13 deletions

identified by GAMToC-L include both RB1, which is known to co-occur in Proneural type

glioblastomas with TP53 mutations, as well as DIS3, an exonuclease.

In contrast to this consistency with GAMToC, GAMToC-L finds different regions on

chromosome 15 and on chromosome 14 from the regions selected in GAMToC’s simulated

annealing result. As GAMToC only used the recurrent alterations as input, and these recurrent

deletions are highly linked to the genes chosen, the recurrent mutations may have only been

chosen in GAMToC for their linkage to these module members. The chromosome 15q deletion,

identified by GAMToC as TMCO5A, is not chosen by GAMToC-L for the module. Instead

GAMToC-L chooses a region containing PIAS1, the protein inhibitor of activated STAT1, LCTL,

and MAP2K5. The deletion that GAMToC selected on chromosome 14 was PTPN21. But

GAMToC-L selects others in the broad peak containing that gene, particularly CDCA4, and

immediately 3’ of CDCA4, a G-protein coupled receptor, GPR132. The gene CDCA4 has been

shown to repress E2F in regulation of cell proliferation(Hayashi et al. 2006).

A number of other copy number alterations also appear to co-occur in the TP53 deleted samples.

A deletion in chromosome 12 containing OSBPL8, Oxysterol binding protein-like 8, is not

significantly recurrent. But GAMToC-L selects it, rather than its recurrently amplified neighbors

Page 70: Rachel D. Melamed

59

on 12q, because of its particularly strong relation with chromosome 11 deletion and TP53

alteration. Although this protein is not well studied in relation to cancer, this alteration may

influence the well known metabolic transformation in cancer cells known as the Warburg

effect(DeBerardinis et al. 2008). In addition, the cholesterol binding activity of other members of

the oxysterol binding protein family has been shown to function as a scaffolding protein

influencing ERK phosphorylation, in a key cell signaling pathway(P.-Y. Wang, Weng, and

Anderson 2005). This is an example of a locus that is not identified by recurrence, but that shows

a strong pattern of relationship with other alterations. Another example is found in chromosome

4. The loci identified there are also not recurrently deleted, but we find NAA15 and FBXW7 have

strong relationships with TP53 and other TP53-co-ocurring genes. The protein N-terminal

acetyltransferase NAA15 may regulate translation and apoptosis(Arnesen et al. 2006). As

mentioned in 2.1.2, FBXW7 is a known tumor suppressor in other cancers, but deletions in

glioblastoma have not been reported.

Deletions in chromosome 6, while not particularly correlated with TP53 alteration, have a

correlation with the TP53-associated GAMToC-L genes. Another set of genes is consistently

anticorrelated with the TP53 group. This set of alterations show some positive correlation

amongst each other. This includes the deletion to the CDKN2A locus, and co-occurring

amplifications in chromosome 19 and chromosome 20. A final locus chosen by GAMToC-L is

chromosome 1 amplifications: these are associated with TP53 and other of TP53's companions,

but also show positive correlation with the chromosome 20 amplifications. These represent an

interesting exception to the overall pattern of two mutually exclusive sets of genetic alterations.

When the results from the first iteration of GAMToC-L are removed, and the algorithm is re-run,

a module similar to the greedy module from GAMToC appears (Figure 3-14), containing co-

Page 71: Rachel D. Melamed

60

occurring ATRX and IDH1 mutations, mutually exclusive with co-occurring chromosome 10

deletion and EGFR amplification. Some additions include PIK3R1, which is known to co-occur

in IDH1-bearing glioblastomas, along with amplification of 8q24, near MYC, that occurs more in

the IDH1 mutant samples.

Figure  3-­‐14:  Second  module  from  GBM  data.  For  legend  see  Figure  3-­‐13.  

 

IDH1 @^2 GTPBP4,LARP4B @^10

EGFR @^7

@8q24

ATRX @^X

@16p

HSPA13 ITSN1 NRIP1 @*21

PIK3R1 @^5

Page 72: Rachel D. Melamed

61

3.2.2.2 Results from lower grade gliomas Lower grade gliomas (LGG) represent an interesting point of comparison with glioblastoma.

Glioblastomas often arise from lower grade gliomas, and all gliomas generally arise from glial

cells. Like glioblastoma, lower grade gliomas have a number of subtypes, linked to distinct

phenotypic outcomes. This is reflected in the top LGG module (Figure 3-15). As is well known,

a strongly co-occurring deletion in 1p and 19q mutation is mutually exclusive with TP53 and

ATRX mutation. Both of these mutually exclusive genomic subtypes frequently have mutation in

IDH1. This pattern is reflected by the module chosen. Additionally, EGFR amplification and

chromosome 10 deletion, lesions similar to the worse-prognosis classical subtype of

glioblastoma, appear mutually exclusive with the IDH1-co-occurring alterations, much like in

glioblastoma. Some nuance is added to this pattern. Again in resemblance to glioblastoma,

chromosome 11 deletions co-occur with TP53 deletion in lower grade gliomas. The genes chosen

include not only BRSK2, but also CDKN1C, a cell cycle regulator. Also co-occurring with the

TP53 group are amplifications affecting DEPTOR, also located near MYC. While MYC

amplification can promote cell cycling, DEPTOR amplification is expected to have a different

oncogenic effect, promoting Akt activation and inhibiting apoptosis (Pei et al.). Deletions

containing NAA15, FBXW7, and other genes in 4q31 seem to co-occur more in the 1p19q cases,

mostly lower grade oligodendrogliomas. A final region that is very interesting is SFI1, which

does not fall into one of the three subtypes. It is mutually exclusive with the 1p19q cases, and co-

occurs with both 11p15 deletion and the chromosome 7 and 10 copy number changes. This gene

appears to be relevant in chromosomal segregation and thus may regulate cell cycle in a variety of

contexts.

Page 73: Rachel D. Melamed

62

Figure  3-­‐15  Module  for  lower  grade  glioma.  For  legend  see  Figure  3-­‐13.  

3.2.3 Discussion Our new method has demonstrated power to explore the landscape of high total correlation gene

modules and find those combinations of genes that have a shared non-random pattern of

alteration. We have turned a limitation of copy number data, the linkage between genes on the

IDH1 @2

@11p15

TP53 @17

ATRX @X

@19q

DEPTOR @8

EGFR @7

GGT5 SFI1 SLC5A4 @*22

@10

@4q31

CDKN2A @9

@1p

Page 74: Rachel D. Melamed

63

same chromosome, into an advantage. The simulated annealing process, run over many iterations,

is an effective competitive procedure in which genes from the same chromosome are more likely

to be retained in modules if they have a stronger total correlation with other module genes. This

allows us to prioritize important copy number changes that may be drivers, and to identify

important genetic alterations without using recurrence.

A few improvements could be envisioned for this approach. In particular, it will be interesting to

further explore the differences in modules containing different genes from the same chromosome.

Genes from the same chromosome will usually have a strong correlation with the same genes, but

more subtle patterns may exist. For example FBXW7 and NAA15 deletions, both on chromosome

4, are both highly correlated with TP53 and RB1 alterations, but there may be some distinction

between the modules containing each of these genes. Methods to dissect these patterns at the sub-

chromosomal level will provide further insight into driver alterations and their strongest

relationships. Another improvement could be some mechanism to weight point mutations such

that they appear in the module data at a similar frequency as copy number changes.

Finally, it will be interesting to apply this method to skin melanoma. Melanoma does not have the

distinctive subtype pattern present in the gliomas, so we would not expect the same pattern of

mutually exclusive sub-modules of genes. Thus, as expected, preliminary results from

GAMToC-L show subsets of co-occurring genetic alterations. As melanoma undergoes a high

rate of genetic damage, a method that can sort through the many passenger alterations and find

subsets of cooperating driver genes will be of high interest, even if distinctive molecular subtypes

are absent.

Page 75: Rachel D. Melamed

64

 4 Genetic similarity between cancers and comorbid Mendelian

diseases identifies candidate driver genes Despite large-scale cancer genomics studies that uncover key somatic mutations driving cancer,

most studies, and much of my dissertation work, focus only on patterns of genomic alterations in

tumors. In this section I propose that analysis of comorbidities of Mendelian diseases with

cancers provides a novel, systematic way to discover new cancer genes. If germline genetic

variation in Mendelian loci predisposes bearers to common cancers, the same loci may harbor

cancer-associated somatic variation. Compilations of clinical records spanning over 100 million

patients provide an unprecedented opportunity to assess clinical associations between Mendelian

diseases and cancers. I systematically compare these comorbidities against recurrent somatic

mutations from more than five thousand patients across many cancers. Using multiple metrics for

genetic similarity, I show that a Mendelian disease and comorbid cancer are indeed have genetic

alterations of significant functional similarity. This result provides a basis to identify candidate

drivers in cancers including melanoma and glioblastoma. Some Mendelian diseases demonstrate

“pan-cancer” comorbidity and shared genetics across cancers. This work is under review at

Nature Communications.

 

Page 76: Rachel D. Melamed

65

 4.1 Introduction

Recent years have brought an explosion in the number of genomically profiled tumors, and the

promise of finding most genetic loci containing cancer-predisposing variation seems within reach.

While algorithms to sort through the complex landscape of tumor lesions(Lawrence et al. 2013;

Mermel et al. 2011) have revealed recurrently altered “driver loci” – those somatic or germline

genetic defects that are most likely to trigger the disease – the directory of relevant genes and the

catalogue of their roles in tumor progression remain incomplete. The search for cancer genes has

expanded to additional informative patterns, such as mutual exclusivity of mutation across

patients and functional relationships between cancer-altered genes(Ciriello et al. 2011; Vandin,

Upfal, and Raphael 2012; G. Wu, Feng, and Stein 2010).

One historical source of information on key cancer alterations may be found in Mendelian

disorders, rare conditions that have long provided insight into a wide array of human disease

processes. Some of the first genes linked to cancer were characterized by their highly penetrant

familial association with certain tumors. Studies of familial retinoblastoma led to the

identification of RB1 as a tumor suppressor(Friend et al. 2014), while cases of Li-Fraumeni

syndrome showed that germline mutation of TP53 pleiotropically predisposes patients to many

cancers(Malkin et al. 1990). Other Mendelian disorders, such as Rubinstein-Taybi syndrome,

involve a primary phenotype apparently unrelated to cancer, yet the bearers are known to have an

increased tumor risk(R. W. Miller and Rubinstein 1995). Recent studies demonstrating that

Rubinstein-Taybi’s primary causative gene, CREBBP, is also recurrently somatically inactivated

in a number of cancers(Pasqualucci et al. 2011; Kishimoto et al. 2005; Yang 2004; Mullighan et

al. 2011) have provided a likely explanation for this comorbidity. These examples suggest that

Mendelian germline mutations could predispose Mendelian disease patients to common cancer by

Page 77: Rachel D. Melamed

66

disrupting cellular functions that in the majority of cancer patients are altered by somatic rather

than germline genetic events.

Recently, Electronic Health Record (EHR) data sets of unprecedented size have provided

statistical power to measure comorbidity of pairs of diseases(Blair et al., 2013; D.-S. Lee et al.,

2008; Park, Lee, Christakis, & Barabási, 2009). With the recent increase in the amount of data

recorded in EHRs it is newly possible to detect clinical associations even in diverse rare diseases,

such as some Mendelian diseases. These results have suggested that comorbidity is indicative of

shared germline genetic architecture. Here, we propose that Mendelian disease comorbidity with

cancer could be associated with a relationship between Mendelian disease loci and driver loci

somatically altered in cancer. It is possible that genetic variants that cause Mendelian disease with

high cancer comorbidity also provide a selective advantage to aberrant cells of a developing

tumor, leading to this predisposition to a certain type of cancer. If this is correct, exactly the same

Mendelian loci and molecular pathways incorporating their products would be involved in a

somatic context in tumors of patients lacking the germline mutation. Thus, comorbidity calculated

from EHRs spanning large numbers of patients could provide a novel line of evidence for

functional involvement of some genes as cancer drivers.

By integrating clinical data from more than 100 million patients with somatic genomic

information from thousands of tumors from The Cancer Genome Atlas (TCGA)(“The Cancer

Genome Atlas”), we explore the connection between Mendelian diseases and common cancers.

First, we examine the hypothesis that comorbidity between Mendelian disease and cancer may be

due to similarities between the genes involved in each. We find that comorbid diseases display

statistically significant genetic similarity. Then, we use this relationship to test genetic similarity

for comorbid pairs of Mendelian disease and cancer, identifying those disease pairs with shared

Page 78: Rachel D. Melamed

67

cellular processes. For each cancer, we prioritize these comorbid and genetically similar

Mendelian disease genes and pathways as candidate novel cancer drivers.

4.2 Comparing Mendelian disease and comorbid cancer

4.2.1 Integration of disease comorbidities and genes In the work of Blair, et al.(Blair et al. 2013) the authors estimated comorbidity for a set of

diseases well characterized by patient billing codes, comprising 95 Mendelian diseases and 65

complex diseases, including 13 common cancers. Comorbidity was calculated using seven EHR

datasets, including the MarketScan insurance claims data covering nearly 100 million patients.

For each complex disease, they compared its incidence in Mendelian disease patients against its

marginal incidence. They crossed patient zip code information with US census data to obtain

demographic, socioeconomic, and environmental factors. Then they corrected for these

confounders, as well as for errors in billing codes, using a regression approach. Combining these

analyses, they estimated relative risk for a complex disease in Mendelian disease patients, as well

as a significance level. We use these estimates throughout this work. For each Mendelian disease

billing code set, the authors curated a list of corresponding diseases, each linked to genetic

loci(McKusick-Nathans Institute of Genetic Medicine, Johns Hopkins University (Baltimore).

We updated the mapping of diagnosis codes to genes using OMIM as well as OrphaNet

data(Hoehndorf, Schofield, and Gkoutos 2013). Utilizing their work and other curation, we find a

median of four genes related to each Mendelian disease type (the full distribution is shown in

Figure 4-1a).

Of the 13 cancer diagnosis code sets included in the Blair analysis, 10 correspond to one or more

tumor types profiled in TCGA. These 10 diagnosis codes correspond to 15 TCGA tumor types,

including melanoma, glioblastoma, and other common cancers, with genomic data across a total

of 5,667 patients. We downloaded the calls of recurrently altered genes as assessed by the Broad

Page 79: Rachel D. Melamed

68

Institute and made available in the Firehose (http://www.broadinstitute.org/cancer/cga/Firehose)

download data set of 9/23/2013. MutSigCV(Lawrence et al. 2013) assigns a statistic for evidence

of selection for mutation of a gene across a set of tumors. For each tumor type, we select those

genes with a q-value statistic less than .25. GISTIC2(Mermel et al. 2011) identifies genes in

significantly recurrent and focal regions of copy number amplification or deletion, and we include

only the genes in copy number peaks that contain fewer than 50 genes. Each tumor type has from

zero to hundreds of associated genes either mutated or copy number altered. A median 155 genes

are recurrently genetically altered per tumor type (Figure 4-1b). Entrez gene info data

(ftp://ftp.ncbi.nlm.nih.gov/gene/DATA/GENE_INFO/Mammalia/Homo_sapiens.gene_info.gz )

was used to find common identifiers between all data sets

Page 80: Rachel D. Melamed

69

 Figure  4-­‐1  Distribution  of  number  of  genes  per  disease.    

a.  The  number  of  genes  per  Mendelian  disease  group  ranges  from  0  (for  chromosomal  disorders)  to  51  (for  retinitis  pigmentosa),  with  a  median  of  4  genes  per  disease.    b.  The  number  of  recurrent  genes  per  cancer  ranges  from  11  for  kidney  chromophobe  to  397  for  lung  adenocarcinoma,  with  a  median  of  155  genes.  

As can be seen in Figure 4-1, some Mendelian diseases have multiple causal genes and the

severity and rarity of Mendelian diseases also varies widely. We investigate factors related to the

number of genes per Mendelian disease, and we find that this number is somewhat associated

with severity and with rareness of the Mendelian disease, two factors that can influence the

overall population of surviving adults with the disease (Figure 4-2a-b).

0 10 20 30 40 500

10

20

30

40

50

60

70

Num genes per Mendelian disease group

Num

Men

delia

n di

seas

e gr

oups

a

b

0 50 100 150 200 250 300 350 4000

1

2

3

4

5

Num genes per cancer

Num

can

cers

Page 81: Rachel D. Melamed

70

 Figure  4-­‐2  Characteristics  of  Mendelian  diseases  

a,b.  Each  point  is  a  gene.  For  each  gene  we  plot  the  number  of  other  genes  associated  with  the  Mendelian  disease  category,  against  the  rareness  and  severity  are  associated  with  the  particular  Mendelian  disease  variant  as  cataloged  in  Orpha  database.  For  example,  the  gene  ZIC2  is  associated  with  holoprosencephaly,  which  has  a  total  of  10  genes,  and  ZIC2's  associated  disease  is  annotated  at  a  frequency  of  1-­‐9  /  100,000  (the  second  least  rare  category).    It  is  important  to  note  that  this  information  is  approximate  and  contains  many  missing  values.  Only  genes  with  available  information  are  plotted.  c.    Each  point  is  a  Mendelian  disease,  and  the  number  of  genes  associated  with  a  disease  is  significantly  correlated  with  the  number  of  cancer  comorbidities  

 Most importantly for this study, we find that number of genes per Mendelian disease is correlated

with cancer comorbidity ((Figure 4-2c). This has a number of possible explanations. One is that

more rare, or more rarely diagnosed, diseases lack power to detect both causal genes and to detect

comorbidities in clinical records. Another explanation is that Mendelian diseases with more

genes annotated are more likely to have disease subtypes (one or more of these causal genes) that

are related to cancer. In any case, any analysis of the comorbidity data must take this association

into account.

# genes per MD

# co

mor

bid

canc

ers

per M

D

Spearman rho = 0.25 p=0.02

10 20 30 40 500

2

4

6

8

10

12

Number of genes in the associated disease

Rar

enes

s ca

tego

ry (5

= m

ost c

omm

on)

Spearman rho = 0.10 p=0.026

10 20 30 40 501

1.5

2

2.5

3

3.5

4

4.5

5

Number of genes in the associated disease

Age

of d

eath

cat

egor

y (5

= n

orm

al)

Spearman rho = 0.16 p=0.005

10 20 30 40 501

1.5

2

2.5

3

3.5

4

4.5

5a b

c

Page 82: Rachel D. Melamed

71

4.2.2 Genetic similarity of comorbid diseases Next, we compare the sets of genes associated with a Mendelian disease to the recurrently

genetically altered genes in TCGA. We consider multiple genetic similarity metrics, with the goal

of assessing whether comorbidity is significantly related to shared genetics. The approach is

outlined in Figure 4-3a.

 Figure  4-­‐3  Outline  of  the  approach.    

a.  Integration  of  the  data  and  overview  of  genetic  similarity  metrics.  b.  Examples  of  comparison  of  pairs  of  diseases.    All  comorbid  pairs  of  Mendelian  disease  and  cancer  with  TCGA  data  are  compared.    Genetic  similarity  of  comorbid  diseases  is  assessed  using  multiple  metrics.  A  simple  combination  of  presence  of  any  one  of  the  genetic  similarity  metrics,  after  correcting  for  the  number  of  comorbid  pairs,  is  used  to  predict  novel  cancer  driver  loci.  

Our similarity metrics are first evaluated on the aggregate of comorbid diseases in order to test

the hypothesis that comorbidity is significantly related to shared genetic factors. Then, we use

analogous tests for the pairs of diseases, in order to identify comorbid Mendelian disease and

cancer with evidence of related gene sets. Below, we describe both uses of each metric.

OMIM$

Germline altered

EHR$comorbidity$of$diagnosis$codes$$

TCGA$

Map codes to diseases

mutated$amplified$

deleted$

Gather genetically altered genes for each disease

Genetic similarity of comorbid diseases

Com

orbi

d p

airs

of M

ende

lian

dise

ase

and

canc

er

a b

Network connections

Gene enrichment Pathway enrichment

Coexpression

Figure-1 (Rabadan)

Mendelian Cancer Gene,

enrich

Pathw

ay

BioGRID

Huma

nNet

Coexpr

.

Candid

ate

Aromatic)amino)acid)metabolism)(pigment) SKCM ✔ ✔

Heart)&)Skeletal))(Rubinstein=Taybi) SKCM ✔ ✔

Heart)&)Skeletal))(Rubinstein=Taybi) GBM ✔ ✔

Heart)&)Skeletal))(Rubinstein=Taybi) LGG ✔ ✔

Holoprosencephaly LGG ✔ ✔

Holoprosencephaly GBM

Diamond=Blackfan GBM ✔ ✔

Diamond=Blackfan GBM ✔ ✔

... ... ...

Page 83: Rachel D. Melamed

72

In the first genetic similarity metric, we examine whether the same genes responsible for a

Mendelian disorder are more likely to be altered in comorbid cancers. For each of 427 pairs of

comorbid Mendelian disease and TCGA cancer, we assess how many genes are shared (Figure

4-4). The gene enrichment metric scores the overlap of the Mendelian disease gene set of size m,

within a cancer gene set of size c. The score assesses whether the number of genes in the overlap

between the two sets is more than expected. For the per-pair score, we use a binomial model with

success probability based on the fraction of all assayed genes that contain variants associated with

the Mendelian disease, and number of trials corresponding to the cancer recurrently mutated gene

set size c, and number of successes corresponding to the size of the overlap, v, between the sets:

Binomial(v, c, !#  !"#"$

).

In all comorbid pairs, 41 genes are shared between the Mendelian causal gene set and the

recurrently somatically altered cancer gene set. We test whether this number of genes shared

across the 427 pairs of Mendelian diseases and comorbid cancers is more than would be expected

at random. Our test uses a simulated convolution of the 427 binomial tests: for each pair, the

binomial model, as before, has a success probability based on the fraction of total genes that are

Mendelian disease genes, and a number of trials based on the number of recurrent cancer genes.

Thus the convoluted distribution can be simulated as:

BinomialSample(𝑐! ,!!

#  !"#"$).!,!∈!"#"$%&'()&$ In other words, the samples from each comorbid

pair are added to generate an expected distribution. The model is simulated 100,000 times to

compare to the observed value. We find that 41 occurs in 2.1% of random trials (Figure 4-5a).

Page 84: Rachel D. Melamed

73

 Figure  4-­‐4  Genes  shared  in  comorbid  diseases    

The  genes  shared  in  comorbid  diseases  is  counted  across  all  pairs.  By  comparing  it  to  a  null  distribution  based  on  number  of  Mendelian  and  cancer  genes,  we  can  assess  if  more  genes  are  shared  than  expected.  Cancer  abbreviations  are  in  from  TCGA.  

The pathway metric utilizes the NCI Pathway Interaction Database and the PharmGKB subsets of

the Consensus Pathway Database (Kamburov et al. 2013) in order to obtain a diverse and non-

Androgen Insensitivity SyndromeCongenital Ectodermal Dysplasia

Congenital Pigmentary AnomaliesDisorders of Aromatic Amino Acid Metabolism

Sensory Retina DystropiesChondrodystrophy

Chronic Progressive External OphthalmoplegiaCongenital Hirschsprung Disease

Congenital HydrocephalusCongenital Hypogammaglobulinemia

ErythromelalgiaFacial and Skull Anomalies

HoloprosencephalyHuntington Disease

Immunodeficiency with Increased IgMOsteogenesis Imperfecta

Specific Nail AnomaliesSystemic Primary Carnitine Deficiency

Anophthalmos/MicropthalmosFamilial Dysautonomia

Non−Specific Autosomal Deletion SyndromesSpecified Anomalies of the Musculoskeletal System

Friedreich AtaxiaGlycogenosis

"Pervasive, Specified Congenital Anomalies"Non−Specified Osteodystrophy

Diamond−Blackfan AnemiaGlucose−6−Phosphate Dehydrogenase Deficiency

Inherited Adrenogenital DisordersCirculating Enzyme Deficiencies

Congenital Disorders of Purine/Pyrimidine MetabolismRetinitis Pigmentosa

"Polycystic Kidney, Autosomal Dominant"Congenital Ichthyosis

Cystic FibrosisDisorders of Copper Metabolism

Severe Combined ImmunodeficiencyDisorders of Urea Cycle Metabolism

Spinocerebellar AtaxiaThalassemia

Dopa−Responsive DystoniaHypopituitarism

Sickle Cell AnemiaCombined Heart and Skeletal Defects

Degenerative Diseases of the Basal GangliaHereditary Hemorrhagic Telangiectasia

Inherited Anomalies of the SkinLong QT Syndrome

Cerebral Degeneration Due to Generalized LipidosesChronic Granulomatous Disease

Disorders of Phosphorous MetabolismDisorders of Straight Chain Amino Acid Metabolism

Genetic Anomalies of LeukocytesHaemophilia

Hereditary Sensory NeuropathyLipoprotein Deficiencies

LGG

GBM

LUAD

LUSC

UC

ECBL

CA

BRC

APR

ADSK

CM

CO

ADR

EAD

STAD

KIR

CKI

RP

KIC

H

0 5 10 150

5

10

15

comorbidcomorbidityundetected

One gene shared

Two genes shared

Page 85: Rachel D. Melamed

74

redundant set of pathways. The set contains 1343 pathways and a total of 4954 genes. We create a

gene list containing the union of all genetically altered cancer genes across all of the cancers

studied, and we remove all pathways with enrichment in this list in order to filter very general

cancer cellular processes. We score strength of the overlap of a cancer gene set within each gene

set associated with each remaining pathway using the same binomial gene enrichment score, then

corrected by the number of pathways with the Benjamini-Hochberg method(Benjamini and

Hochberg 1995). Many pathways have no overlap with a cancer’s gene list, so the enrichment

score for these is 1.0. For the Mendelian diseases, we consider a pathway to be affected if it

contains any Mendelian disease gene. To assess the similarity for a pair of diseases, we use the

Spearman correlation coefficient of the pathway scores for each disease across all pathways, with

the Spearman significance statistic providing our per-pair score.

For the aggregate score across comorbid pairs, we use a cutoff on cancer enrichment (q-value <

.1), and we count the number of out of the n pathways that are both enriched in the cancers (c),

and involved in the Mendelian disease (m). We find 136 pathways shared in comorbid pairs. We

assess whether this number of overlapping pathways is more than expected using the convolution

of hypergeometrics, similar to the gene enrichment convolution:

HypergeometricSample(𝑛, 𝑐! ,𝑚!).!,!∈!"#"$%&'()&$ The results are shown in Figure 4-5b. In

order to ensure that the significance is not only due to two Mendelian disorders with the most

pathways impacted, we also run this test when Rubinstein-Taybi syndrome and Pervasive

Specified Congenital Anomalies are removed: in this case only 81 pathways are shared but the

overlap is still highly significant.

Our next test of genetic similarity between comorbid diseases uses well-studied gene interaction

networks. The network metric measures the number of direct interactions of each Mendelian

Page 86: Rachel D. Melamed

75

disease gene set with the cancer gene set. This number is compared to the number found in a set

of shuffled networks, created using a degree-preserving randomization algorithm(Maslov and

Sneppen 2002). In this randomization algorithm, a network is shuffled by repeatedly re-wiring

pairs of edges, in order to preserve each node’s number of connections but randomize which

genes are connected to each other. A pair of diseases is considered similar if fewer than 5% of

random networks have the same or higher number of interactions. For the aggregate score, we

count over the Mendelian diseases, the number of edges between a Mendelian disease's genes and

the set of comorbid cancer genes. This count is compared against the count from the shuffled

networks. We use two networks to independently score our disease pairs. In the BioGRID binary

interaction data set(Stark et al. 2006), a curated set of genetic interations and protein interactions,

there are 140,402 edges on 14,112 nodes, covering 86% of Mendelian disease genes and all but

four of our Mendelian disease sets. In all, there are 797 direct edges between comorbid genes in

this network, a number found in less than 2% of random networks (Figure 4-5c). Another

network, HumanNet, is constructed by integrating a number of data sources(I. Lee et al. 2011).

HumanNet trains its integrated data set on Gene Ontology categories of genes, and it assigns a

confidence score, in terms of log-likelihood of interactions, to each learned edge. We take the top

10% most confident edges, resulting in a network with 7,931 nodes and 47,934 edges. In

HumanNet, there are 296 direct edges between comorbid disease genes, which is a number found

in only 0.2% of random networks (Figure 4-5d).

 

Page 87: Rachel D. Melamed

76

Figure  4-­‐5  Aggregate  similarity  of  comorbid  diseases    

a.  The  number  of  genes  shared  among  pairs  of  comorbid  diseases  is  41,  more  than  all  but  2.1%  of  the  generated  null  model  (see  Method  for  details  about  null  model  for  gene  and  pathways  shared).    b.  The  number  of  pathways  shared  is  136.  c,  d.  The  number  of  edges  shared  between  a  Mendelian  disease’s  genes  and  the  genes  involved  in  comorbid  cancers,  shown  for  two  different  gene  networks.  In  BioGRID,  there  are  797  edges,  while  in  HumanNet  296  edges  are  found.    

It is important to note that well-known Mendelian cancer syndromes were removed from the

analysis before testing the association of comorbidity and genetic similarity. This includes: Li-

Fraumeni (TP53, CDKN2A), neurofibromatosis (NF1, NF2), Cowden syndrome and related

hamartomas (PTEN, STK11), tuberous sclerosis (TSC1, TSC2), and dyskeratosis syndromes

(TERT and other genes involved in telomere maintenance). We do not include these known

germline cancer genes in our analysis because we wish to assess the significance of novel

Mendelian disease associations with cancer. These cancer syndromes, as would be expected, are

a

d c

b

10 20 30 40 500

0.1

0.2

0.3

0.4

2.1%

number

rand

om d

istrib

utio

n

Genes shared

40 60 80 100 120 1400

0.1

0.2

0.3

0.4

0%

number

rand

om d

istrib

utio

n

Pathways shared

700 750 800 8500

0.1

0.2

0.3

0.4

1.7%

number

rand

om d

istrib

utio

n

BioGRID edges

220 240 260 280 3000

0.05

0.1

0.15

0.2

0.25

0.2%

number

rand

om d

istrib

utio

n

HumanNet edges

Page 88: Rachel D. Melamed

77

each comorbid with multiple cancers, and they show many shared genes and pathways with the

cancers (Supplementary Table 3).

 Thus, using a number of lines of evidence, we have shown that the genes involved in Mendelian

diseases have a specific functional relationship with the genes altered in co-occurring cancers,

and most of these connections are novel. Therefore, comorbidity may be due to the genetic

similarity relationship. In order to use comorbidity as a source of candidate novel drivers for each

cancer, we use the per-pair scores of genetic similarity that we can apply to each pair of comorbid

diseases. These per-pair metrics are related to the aggregate measures, as discussed above.

 To these pairwise and aggregate measures of similarity, we wished to add an entirely unbiased

source of information on functional similarity and cell-specific expression. We developed a

coexpression metric utilizing the data from FANTOM5(Consortium, Pmi, and Dgt 2014). The

FANTOM5 data covers a diverse range of 889 cellular states, assessing promoter activity in each

gene in each cell or tissue type. We download the human CAGE peak data quantified by

transcripts per million

(http://fantom.gsc.riken.jp/5/datafiles/latest/extra/CAGE_peaks/hg19.cage_peak_tpm_an

n.osc.txt.gz). Adding all peaks that are assigned to the same gene, we create an estimate of

aggregate expression of each gene in each sample. As we wish to measure whether genes

involved in a pair of diseases are expressed in the same conditions, we calculate coexpression of

pairs of genes using the Pearson correlation coefficient. To calculate our coexpression similarity

for a pair of Mendelian disease and cancer, we consider that significantly elevated coexpression

between any cancer gene and a set of Mendelian disease genes represents interesting similarity.

Thus, for each cancer gene we compare whether the set of Mendelian disease genes has high

coexpression with that cancer gene, as compared against the distribution of coexpression of all

other genes with the cancer gene. We test this for each cancer gene using the Wilcoxon rank-sum

Page 89: Rachel D. Melamed

78

test. The p-values are then corrected for the number of cancer genes tested using the Benjamini-

Hochberg method.

 We correct the genetic similarity scores for the number of comorbid diseases considered.

Coexpression has the most instances of similarity, most likely due to the fact that more genes can

be compared and many types of functional relationships can be captured with coexpression. After

correction, the gene enrichment and network metrics have very few instances of significant

similarity, a point that is discussed below. Our candidates comprise the significantly functionally

connected genes from comorbid and genetically similar disease pairs. The scores are shown in

Supplementary Table 3.

4.3 Mendelian disease comorbidity and cancer processes

4.3.1 Prediction of diseases with shared cellular processes From the per-pair genetic similarity metrics, we have generated a list of candidate linked

Mendelian diseases, and associated genes and processes, across 15 TCGA cancers. The complete

resulting list of genes, and genetic similarity scores, associated with each linked disease pair is

available in Supplementary Table 3. To provide examples and to demonstrate their relevance we

highlight some candidates implicated for cutaneous melanoma and brain neoplasms.

Cutaneous melanoma is often located on sun-exposed sites, undergoing a high rate of genetic

damage. Our findings can highlight both recurrently altered genes in melanoma and comorbid

Mendelian genes as potential cancer drivers. A central transcription factor involved in melanocyte

cell fate, MITF, is related to multiple Mendelian diseases comorbid with melanoma. This gene

has a complex role in this cancer: while it is recurrently amplified in 26% of TCGA melanomas,

possibly promoting melanocyte proliferation, it is also frequently deleted (11% of cases).

Suppression of the gene is also advantageous for the growing cancer, as it reduces terminal

Page 90: Rachel D. Melamed

79

differentiation and senescence in melanocytes(Levy, Khaled, and Fisher 2006; Yajima et al.

2011). The melanocyte’s primary receptor MC1R, upstream of MITF, its other upstream

activators, PAX3 and SOX10, as well as MITF’s key target, TYR, are all associated with

Mendelian disorders comorbid with melanoma (Figure 4-6).

 Figure  4-­‐6  Depiction  of  comorbid  diseases  with  skin  melanoma  

Comorbid  diseases  are  shown  in  terms  of  the  roles  of  the  Mendelian  disease  genes  in  the  melanocyte  development  program  as  well  as  other  cancer  related  processes.    Genes  that  are  recurrently  somatically  mutated  in  melanoma  are  highlighted.    Solid  edges  represent  interactions  from  the  literature,  while  the  dashed  edges  represent  significant  coexpression.    Orange  outlines  represent  genes  with  common  polymorphisms  conferring  increased  melanoma  risk.  

Of these, MC1R and TYR are associated with oculocutaneous albinism (included in International

Classification of Disease, revision 10 (ICD10) billing code E70.2/3, melanoma relative risk 95%

confidence interval (CI) = (2.16 - 5.19)). MC1R is among the recurrently deleted genes in

melanoma. Germline variants of MC1R, causing red hair, have been implicated as a risk factor for

melanoma via both pigmentary and non-pigmentary pathways(Cao et al. 2013; Raimondi et al.

2008), and polymorphic variants of TYR, which leads to a green eyes, also confer significant,

though lesser, risk(Gudbjartsson et al. 2008). Other albinism-related genes have significantly

Q87.2 incl. Rubinstein-Taybi

Q11 microphthalmos

Q79.8 incl. Waardenburg

E70.2/3 incl. albinism

MITF%

MC1R%cAMP

TYR%

OCA2%

SLC45A2%

TYRP1%

PAX3%SOX10%

SNAI2%

SOX2%

OTX2%

PAX6%

BCOR%

TP53%EP300%CREBBP%

Page 91: Rachel D. Melamed

80

elevated coexpression with MITF (p = .020) as well as MITF's target gene(Hoek et al. 2008)

KCNAB2 (p = 0.0093). KCNAB2 is recurrently deleted in the melanoma cases.

While the candidate melanoma genes associated with albinism are not recurrently genetically

mutated in melanoma, we examine their patterns of expression for evidence of a functional

contribution to the disease. We download Level 3 RNASeq data from TCGA portal, and the

RSEM(B. Li and Dewey 2011) expected counts are rounded to create the input to the analysis.

We transform these using the variance stabilizing transformation from DESeq2(Love, Huber, and

Anders 2014), which is recommended for clustering data. We then cluster melanoma tumors by

their expression of these genes using consensus clustering methods implemented in

ConsensusClusterPlus(Wilkerson and Hayes 2010), and we find stable clusters (Figure 4-7a). An

optimum clustering is found (based on change in classification consistency) of k=4. Three main

large clusters are consistent through k=3 to k=6. We assess clinical outcome in these groupings,

using the R package Survival(Therneau 2012) to assess survival difference between the groups

and to plot, based on the available TCGA clinical data. Cluster assignments are highly predictive

of patient survival (p = 0.0022, Figure 4-7b). This suggests that indeed this pathway is highly

relevant for melanoma progression.

Page 92: Rachel D. Melamed

81

 Figure  4-­‐7  Analysis  of  the  role  of  albinism  related  genes  in  melanoma.    

a.  Transformed  expression  of  the  genes  in    471  melanoma  samples  is  shown.    Samples  are  arranged  according  to  their  consensus  cluster  tree,  and  genes  are  clustered  using  Pearson  correlation  coefficient.  The  consensus  cluster  assignments  for  k=4  are  shown  by  the  colored  label  at  the  top.  b.  Survival  analysis  for  the  four  classes  using  the  log  rank  test  shows  a  significant  distinction  in  prognosis  among  groupings  assigned  using  only  the  expression  of  these  genes  

Also regulating MITF activity are its coactivators EP300 and CREBBP(Sato et al. 1997), genes

associated with the melanoma-comorbid Rubinstein-Taybi syndrome (code group Q87.2, relative

Consensus clusters a

0 2000 4000 6000 8000 10000

0.0

0.2

0.4

0.6

0.8

1.0

Chisq= 14.6 on 3 degrees of freedom, p= 0.00215

days

survival

1234

b

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

AP3B1

HPS5

TAT

HPD

FAH

OCA2

BLOC1S3

HPS6

DTNBP1

HPS1

TYRP1

SLC45A2

TYR

5 10 15

Value

0200

400

Color Keyand Histogram

Count

Page 93: Rachel D. Melamed

82

risk 95% CI = (1.19 - 1.99)). EP300 is recurrently amplified (36% of the TCGA melanomas), but

also frequently deleted (7% of cases). Rubinstein-Taybi shares many pathway enrichments in

common with melanoma (Figure 4-8), including "melanocyte development and pigmentation

pathway" and "Regulation of nuclear beta catenin signaling and target gene transcription", both of

which involve MITF. The amplifications of EP300 are significantly more likely to co-occur in the

same patients with MITF amplifications (one-tailed Fisher’s exact test, p = 0.0041), suggesting

cooperation between the alterations, and a particular role for these genes in melanoma: the

histone acetyltransferase activity of EP300 might enhance the function of an oncogenically

amplified MITF. CREBBP and EP300 defects have also been linked to aberrant TP53 and BCL6

regulation in some lymphomas(Pasqualucci et al. 2011).

 Figure  4-­‐8  Pairwise  pathway  metric  for  Rubinstein-­‐Taybi  and  melanoma  

For  a  disease  pair,  the  pathway  metric  compares  the  pathways  impacted  by  the  Mendelian  disease  to  the  pathways  enriched  for  the  cancer  gene  sets.  Here,  pathway  enrichments  for  melanoma  genes  (blue)  are  compared  to  pathways  involved  in  Rubinstein-­‐Taybi  syndrome.  Each  vertical  red  line  represents  one  pathway  impacted  by  a  Rubinstein-­‐Taybi  gene.  The  pathways  are  sorted  by  their  enrichment  in  melanoma.  The  Spearman  correlation  between  the  corrected  p-­‐values  of  melanoma  and  the  impacted  pathways  of  Rubinstein-­‐Taybi  for  the  pathways  is  -­‐.25,  p  =  6.3x10-­‐21.  

Comorbidity of melanoma with ectodermal dysplasias (ICD10 code Q81, melanoma relative risk

95% CI = (6.01-17.84)) may highlight the importance of tissue invasion in melanoma

progression. The ectodermal dysplasia disease epidermolysis bullosa can arise from genetic

alteration to proteins involved in structural support, tissue integrity, and adhesion in the dermis

0 500 10000

0.2

0.4

0.6

0.8

Pathway index (ordered by SKCM enrichment)

Can

cer e

nric

hmen

t (−l

og10

)

SKCMRubinstein−Taybi

Page 94: Rachel D. Melamed

83

and epidermis. Although the chronic inflammation and tissue damage associated with

epidermolysis bullosa may play a role in its known risk for skin cancers, subtypes of the

condition have been shown to lead to skin squamous cell carcinoma that is more aggressive than

in other conditions involving chronic skin scarring(Fine et al. 2009). The ectodermal dysplasia

genes show high coexpression with a few melanoma-altered genes related to cell contact in the

epithelium, especially PTK6 (Figure 4-9).

Figure  4-­‐9  Coexpression  of  ectodermal  dysplasia  genes  with  PTK6  

PTK6    is  a  recurrently  amplified  gene  in  melanoma,.  Its  coexpression  with    all  genes  is  compared  against  its  coexpression  with  the  genes  associated  with  the  comorbid  disease  set  ectodermal  dysplasias  including  epidermolysis  bullosa.    Outliers  are  removed.  The  two-­‐tailed  rank-­‐sum  p-­‐value,  controlled  for  number  of  cancer  genes,  is  2.0x10-­‐6.  

The gene PTK6 is focally amplified in 44% of melanomas and has an identified role in epithelial

invasion and mesenchymal transition in prostate and breast cancers(Brauer and Tyner 2010;

Zheng et al. 2013), but the gene has been rarely studied in melanoma. The TCGA melanoma

cohort is primarily composed of metastasis samples, but the expression data also includes 103

primary tumors, mostly stage IIC, along with 368 metastases. As changes in cell contact and

mesenchymal transition may be related to metastasis state, we compare expression in primary

versus metastasis.

−0.2

0

0.2

0.4

0.6

0.8

1

All genes EctodermalPTK6

Pea

rson

coe

xpre

ssio

n

Page 95: Rachel D. Melamed

84

We use TCGA barcodes (01 for primary tumor, and 06 or 07 for metastasis), to identify the

metastasis and primary samples. We use edgeR to calculate library size factors and estimate

dispersion, followed by assessment of differential expression. We find that PTK6 is significantly

differentially expressed (adjusted p-value = 3.29x10-28). Then, we examine whether the set of

ectodermal dysplasia genes show differential expression, using voom (Law et al. 2014) to

transform the data, allowing use of the camera gene set score(D. Wu and Smyth 2012).

Additionally, we use the limma(Smyth 2004) differential expression t-statistic to form a pre-

ranked input to GSEA(Subramanian et al. 2005) for gene set differential expression analysis. Of

11 ectodermal dysplasia candidate melanoma genes, nine are significantly downregulated in

metastases as compared to primary (gene set differential expression camera p-value = 0.00032,

GSEA p-value = 0, Figure 4-10).

 Figure  4-­‐10  GSEA  plot  of  the  ectodermal  dysplasia  candidates  

Differential  expression  is  in  primary  (upregulated)  versus  metastasis  samples.      

Page 96: Rachel D. Melamed

85

The other cancers included in our study also have informative genetic and clinical links with

Mendelian disease. Diamond-Blackfan anemia, a blood disorder, is comorbid with the brain

neoplasms (ICD10 D61.01, relative risk 95% CI = (9.22 - 28.67)). Indeed, Diamond-Blackfan

patients have risk for seven of the cancer ICD-9 code groups, along with other blood and solid

cancers(Vlachos et al. 2012). Among Diamond-Blackfan’s causal genes is RPL5, a gene that is

significantly deleted in 8% of TCGA glioblastoma and that suppresses MDM2(Dai and Lu 2004)

(Figure 4-11a). MDM2 is recurrently amplified in 15% of TCGA glioblastoma cases. It is an

established oncogene that negatively regulates TP53(Manfredi 2010). Like RPL5, other

Diamond-Blackfan genes RPL11 and RPS7 repress MDM2 in response to ribosomal

stress(Manfredi 2010). The deletion of RPL5 is mutually exclusive with amplification of MDM2

(p=0.033, Figure 4-11b), supporting the role of RPL5 deletion as an alternative mode of TP53

abrogation. While RPL11 is less frequently deleted, it also has a mutually exclusive pattern with

MDM2 amplification (p=0.042). The role of these ribosomal proteins in glioblastoma appears to

be unstudied, making this an exciting novel finding.

 Figure  4-­‐11  Interaction  of  Diamond-­‐Blackfan  anemia  genes  with  glioblastoma  altered  genes.    

a.  Summary  of  genes  and  their  known  interactions.  b.  Summary  of  copy  number  changes  to  MDM2  and  the  Diamond-­‐Blackfan  associated  ribosomal  proteins  known  to  suppress  the  action  of  MDM2.  Among  the  ribosomal  genes,  RPL5  is  recurrently  and  focally  deleted  such  as  to  be  in  the  GISTIC2  results,  and  it  shows  mutual  exclusivity  with  MDM2  amplification.  RPL11  deletion  is  less  frequent  but  it  is  also  mutually  exclusive.  RPS7  and  RPL11  deletions,  together  with  RPL5  deletions,  form  a  weaker  mutually  exclusive  trend  with  MDM2  (p  =  0.060).  

D61.01 Diamond-Blackfan

RPS7% MDM2%

TP53%RPL5%RPL11%

a

561 GBM copy number profiles

RPL11 RPS7 RPL5 MDM2

b

Page 97: Rachel D. Melamed

86

While Diamond-Blackfan anemia is comorbid with many cancers, the cranial development

disorder holoprosencephaly is only comorbid with the brain neoplasms (ICD10 Q04.2, relative

risk 95% CI = (9.30 - 15.95)). Defects in genes that regulate cranial-specific components of the

sonic hedgehog pathway are responsible for the improper embryonic patterning in

holoprosencephalies(Taniguchi et al. 2012). This pathway regulates expression of the GLI

transcription factors, which have been linked to maintenance of stemness in gliomas(Clement et

al. 2007). Subtypes of glioblastoma have been defined on the basis of gene expression patterns,

and among these the Classical subtype has a signature including Sonic hedgehog

signaling(Verhaak et al., 2010a). Holoprosencephaly genes have weak pathway enrichment

similarity with low-grade glioma genes, as well as coexpression with multiple of the low-grade

glioma genes, particularly the recurrently copy number altered gene VENTX (p = 0.0092). In the

TCGA lower grade glioma cohort, VENTX lesion occurs more in higher grade tumors, and these

lesions are anticorrelated with IDH1 mutation. Mutation of IDH1 is associated with good

prognosis and particularly co-occurs in subtype of low grade glioma with either TP53 alteration

or 1p19q codeletion(Bourne and Schiff 2010). Comparing the IDH1 mutated against the VENTX

mutated samples for patients with both mutation and expression data available, we find strong

differential expression of the holoprosencephaly genes TGIF1, SIX3, ZIC2, GLI2. We use the

same methods as detailed previously to assess differential expression of the set of genes. As a set,

the holoprosencephaly candidate brain neoplasm genes are significantly upregulated in the

VENTX mutated tumors (camera p-value = 0.048, GSEA p-value = 0.031,Figure 4-12). Both

VENTX mutation and activated hedgehog signaling are thus associated with higher grade gliomas.

Changes in regulation of the sonic hedgehog pathway may be an important step in the progression

of lower grade glioma, as is known to be true in the Classical subtype of glioblastoma.

Page 98: Rachel D. Melamed

87

 Figure  4-­‐12  GSEA  plot  of  holoprosencephaly  candidate  genes  

Differential  expression  of  these  genes  is  compared  in  the  VENTX  copy  number  altered  samples  (upregulated)  versus  the  IDH1  mutated  samples.    The  hedgehog  related  genes  are  upregulated  in  the  VENTX  altered  samples.  

4.3.2 Pan-cancer Mendelian associations Above, we describe a number of processes aberrantly regulated in Mendelian disease and in

common cancer. The Blair analysis(Blair et al. 2013) suggested that the unique set of Mendelian

diseases comorbid with a complex disease represented a sort of barcode, indicative of the unique

set of cellular processes underlying each disease. This hypothesis indeed is reflected in the sets of

disorders, and underlying genetic lesions, found in this study.

On the other hand, some Mendelian diseases predispose carriers to many cancer types, while

others have no relationship with cancer. In fact, the number of comorbid cancers per Mendelian

Page 99: Rachel D. Melamed

88

disease follows a highly non-random distribution (Figure 4-13).

Figure  4-­‐13  The  distribution  of  the  number  of  comorbid  cancer  diagnosis  codes  per  Mendelian  disease  

The  actual  distribution  (red  bars)  includes  a  large  number  of  Mendelian  diseases  with  no  cancer  relationship,  and  a  long  tail  with  Mendelian  diseases  that  are  comorbid  with  many  cancers.  The  blue  bars  represent  the  expected  distribution:  about  one-­‐third  of  the  pairs  of  disease  have  a  comorbidity  relationship,  thus  the  expected  mode  of  the  distribution  would  have  four  comorbid  cancers  per  Mendelian  disease.    The  expected  distribution  is  modeled  using  a  binomial.  

One interpretation of this pattern is that the genes altered in some Mendelian diseases, such as Li-

Fraumeni syndrome, Rubinstein-Taybi syndrome, and Diamond-Blackfan anemia, are related to

pan-cancer processes common to cancer development in many contexts. This interpretation is

supported foremost by our finding of statistically significant genetic similarity in comorbid

disease pairs. Additionally, we examine four new cancers with available TCGA data but no

comorbidity information. If the pan-cancer Mendelian diseases impact core cancer processes, we

would expect these to be relevant to these new cancers. We test whether pathways associated with

Mendelian diseases with many (more than five) cancer comorbidities are enriched in the four new

cancers. We find that the Mendelian diseases with multiple comorbidities share 20 pathways with

the four cancers with no comorbidity information, more than the random expectation (p = 0.051,

excluding Mendelian cancer syndromes). In another test of this hypothesis, we assess whether

Mendelian diseases with more cancer comorbidities are associated with genes that have cancer-

0 2 4 6 8 10 120

0.05

0.1

0.15

0.2

0.25

0.3

Number of cancer comorbidities per Mendelian Disease

Frac

tion

of M

ende

lian

dise

ases

expected at randomobserved

Page 100: Rachel D. Melamed

89

related characteristics. We create a set of the 48 genes recurrently altered in more than four of the

19 TCGA tumor types. We call these the multi-cancer mutation genes. Examining FANTOM5

coexpression of the Mendelian disease genes and the multi-cancer mutation genes, and we find a

significant correlation with number of cancer comorbidities in the gene's associated Mendelian

disease. That is: the more cancers that are comorbid with a Mendelian disease, the higher is the

coexpression of a Mendelian disease gene and multi-cancer mutation genes (Spearman

correlation p-value = 0.027). These findings suggest that some Mendelian diseases predispose

patients to many cancers by genetic alteration affecting pan-cancer processes.

Page 101: Rachel D. Melamed

90

The Mendelian diseases with the most links to cancer indeed impact pathways shared across

many cancers, including telomere maintenance, DNA damage response, and mTOR signaling

(Figure 4-14, and Supplementary Table 3).

Figure  4-­‐14  Mendelian  diseases  with  broad  cancer  links  

Those  Menndelian  diseases  that  have  comorbidity  with  and  genetic  similarity  to  more  than  3  cancers  are  compared  to  all  19  available  TCGA  cancers,  15  of  which  have  comorbidity  information.    These  mostly  have  widespread  comorbidity  and  show  genetic  similarity  (after  multiple  testing  correction)  across  many  cancers.  Similarity  was  calculated  here  without  removing  the  known  germline-­‐associated  cancer  genes  in  order  to  view  all  associations.  

Pan-cancer associations with immunodeficiency syndromes could be due to the compromised

immune system, rather than the ability of the tumor to evade immune suppression. However, we

find many instances of genetic similarity with cancer, suggesting that the same functions are

"Polycystic Kidney, Autosomal Dominant"

Congenital Ichthyosis

Severe Combined Immunodeficiency

Inherited Anomalies of the Skin

Hypopituitarism

Spinocerebellar Ataxia

Disorders of Urea Cycle Metabolism

Neurofibromatosis

Specified Hamartoses

Chronic Granulomatous Disease

Lipoprotein Deficiencies

Hereditary Sensory Neuropathy

Combined Heart and Skeletal Defects

Li Fraumeni and Related Syndromes

Retinitis Pigmentosa

Diamond−Blackfan Anemia

Tuberous Sclerosis

PRAD

KIR

C

KIR

P

KIC

H

BLC

A

LUAD

LUSC

BRC

A

LGG

GBM

SKC

M

REA

D

CO

AD

STAD

UC

EC

THC

A

OV

HN

SC

LAM

L

0 5 10 150

5

10

15

comorbid

no comorbidity

unmeasured

Coexpression metric

Pathway metric

BioGRID metric

HumanNet metric

Gene enrichment

Page 102: Rachel D. Melamed

91

frequently somatically altered in tumors. For example, the gene B2M is recurrently mutated or

deleted in the TCGA melanoma, lung squamous cell carcinoma, and colon adenocarcinoma. Loss

of this gene leads to abolition of the MHC class I complex in tumor cells and has been shown to

influence immune escape in some lymphomas(Challa-Malladi et al. 2011). B2M has significant

coexpression with the immunodeficiency genes, and CIITA and RFX5, immunodeficiency genes

that mainly regulate MHC class II expression, have a secondary role in regulating MHC class I

expression(Kobayashi and van den Elsen 2012). Novel pan-cancer associations include the set of

lipoprotein deficiencies, defects in widely expressed proteins that lead to imbalance of blood

cholesterols. The genes associated with lipoprotein deficiencies also influence inflammation and

are enriched in the highly cancer-relevant TGF-β pathway. Cancers, with their elevated rates of

proliferation, are thought to have high cholesterol metabolism, and the role of blood cholesterol in

tumor progression is a current area of research(Llaverias et al. 2011). The lipoprotein deficiency

genes are significantly coexpressed with a number of metabolism related genes that are

recurrently mutated in multiple cancers (Supplementary Table 3). These include IDH1, a gene

that has been shown to be regulated with cholesterol levels(Shechter et al. 2003) and to be

relevant in gliomas and other cancers(Turcan et al. 2012). If pan-cancer Mendelian associations

exist, this further supports the hypothesis that comorbidity between Mendelian disease and cancer

is due to shared processes disrupted by germline or somatic alterations, respectively.

4.4 Discussion

We have shown that Mendelian diseases that are comorbid with a cancer are likely to involve

mutation of genes similar to those that are somatically altered in that cancer. Importantly, this

suggests that comorbidity between Mendelian disease and cancer may be due to germline

mutations that provide a fertile ground for growth of certain aberrant cells. This novel finding

provides new insight into the somatic genetic alterations present in a cancer, presenting them in

the context of well-characterized diseases with simpler genetics. While algorithms for classifying

Page 103: Rachel D. Melamed

92

genes as preferentially somatically mutated in a cancer are an active area of research, comorbidity

can provide an orthogonal line of evidence for involvement of some cellular processes in

oncogenesis and pinpoint driver genes among the recurrently mutated genes. Candidate drivers

among the Mendelian disease genes include many genes that are less recurrently somatically

mutated, but that impact the same pathways. Many of our candidate drivers have a bulk of

evidence supporting their role: beyond our findings related to comorbidity and genetic similarity,

the candidate genes include some recurrently mutated in cancer, and some with identified roles as

drivers in other tumors. Additionally, we have used patterns of co-occurrence of candidate

mutations across tumor cohorts to demonstrate a likely role for these genes in the tumors. For less

frequently mutated candidate drivers, we have related gene expression with clinical indicators.

Our results are informative of the many processes that are involved in cancer development.

Inactivation of ribosomal protein RPL5, associated with Diamond-Blackfan anemia, has the

potential to cause aberrant TP53 degradation in multiple cancers. As cancer is known to involve

defects in differentiation(Hanahan and Weinberg 2011), much like a number of Mendelian

diseases, a role for the Mendelian disease genes in cancer dedifferentiation and aberrant

proliferation is plausible. Other “hallmarks of cancer”, such as invasion or regulation of

apoptosis are also represented in the Mendelian diseases. As cancers have many altered processes

in common, it is logical that we also find some “pan-cancer” Mendelian diseases with multiple

genetic and clinical associations.

In contrast, some germline variants predispose patients to a more narrow range of cancers, which

can reveal more specific oncogenic processes. A few Mendelian disorders are comorbid only with

brain neoplasms and melanoma. As melanocytes are descended from the neural crest, Mendelian

genetic lesions affecting neural development are likely to affect processes in melanocytes,

Page 104: Rachel D. Melamed

93

including proliferation and terminal post-mitotic differentiation. One interesting example is

microphthalmos, meaning small eye, a disease phenotype that, in the mouse, gave rise to the

name of the melanoma oncogene MITF (microphthalmos transcription factor). In humans, the

most common causal genes are closely tied in expression and in function to MITF(Adameyko et

al. 2012) (Figure 4-6 ). Some of the microphthalmos genes have been implicated in neural

derived tumors(Bunt et al. 2010; C. G. Li and Eccles 2012; Yamamoto, Abe, and Emi 2014), and

these may be exciting novel candidates in melanoma. There is a link between some sensineural

disorders and pigment anomalies: the phenotype of microphthalmos can also occur to varying

degrees in patients with Rubinstein-Taybi syndrome and in patients with Waardenburg syndrome,

a pigment and deafness disorder. The idea that disorders comorbid with the same cancer may

share pathways with each other is highly intriguing. Waardenburg syndrome (included in ICD10

code group Q79.8), like microphthalmos, shows comorbidity only with melanoma and brain

neoplasms. Waardenburg has correlated pathway enrichment to melanoma (p = 5.8x10-4): both

diseases are impact melanocyte development and β-catenin signaling pathways. However, the

billing code used is not specific enough to have significant enrichment.

In fact, many of the Mendelian diseases with an apparent risk for cancer do not display genetic

similarity by our metrics. We chose a limited number of genetic similarity metrics in order to

consider different lines of interpretable evidence for functional similarity, but other comparisons

of genetic similarity could capture more connections. For example, the blood disorder thalassemia

can lead to overloaded blood iron levels(Tanno et al. 2007) which may explain these patients’ risk

for a variety of cancers(Torti and Torti 2013); however, this effect is not detected by our current

approach. Additionally, a number of factors introduce noise into our source data. These issues

include ambiguity of the diagnosis codes; heterogeneity of the Mendelian diseases; insufficient

sampling of the mutation spectrum of both Mendelian disease and of cancer.

Page 105: Rachel D. Melamed

94

Our finding of statistically significant association of genetic similarity with comorbidity, despite

these factors, is a main discovery of our work. This implies that future large scale studies mining

rich data sources such as the eMERGE network(McCarty et al. 2011) will find more genetic and

clinical associations. Other future work building on our results includes, foremost, the

experimental assessment of the novel candidate driver genes. Drugs that target these cellular

processes, perhaps as studied in the Mendelian disease patients, may be applicable for the

treatment of the tumors(Brinkman et al. 2006).

Page 106: Rachel D. Melamed

95

5 Data-driven discovery of seasonally linked diseases from an Electronic Health Records system

The Electronic Health Record (EHR) is a rich source of data on patterns of human disease. Health

records include free text entries as well as coded terms, such as the diagnosis coding system ICD-

9-CM (International Classification of Diseases, Ninth Revision, Clinical Modification). Finding

significant disease comorbidities using ICD-9 codes has had implications for the underlying

genetic factors of some diseases (Blair et al. 2013) and has been used to suggest unforeseen

causes or consequences of disease (Holmes et al. 2011). In another section of the dissertation, I

discuss how coded data from clinical records can be combined with cancer genomic information

to better understand cellular processes in cancer. In this section of my dissertation I describe an

exploratory method to use ICD-9 data to detect seasonal patterns in human disease. Temporal

patterns in human disease often reflect changing environmental factors, as is evident in levels of

allergic disease in spring and fall, vector-borne and enteric diseases in summer, and respiratory

infectious diseases in winter. Thus, discovering temporal associations can potentially inform us of

unconsidered causes of a wide variety of human diseases. As EHRs increasingly compile clinical

information from large numbers of patients in a computationally accessible form, they represent a

unique opportunity to seek these patterns. When this data is explored with appropriate methods,

unbiased discovery of trends in incidence could illuminate a diverse array of health conditions.

Additionally, as ICD-9 is an international standard, a uniform methodology could potentially be

applied across EHR data from multiple systems. My goal is to examine properties of temporal

patterns as observed in ICD-9 codes from the EHR and to relate the discovered patterns to the

biology of disease development. This chapter mostly consists of work that was published in

(Melamed, Khiabanian, and Rabadan 2014).

 

Page 107: Rachel D. Melamed

96

5.1 Introduction

Seasonal disease incidence is often associated with environmental or behavioral risk factors for

these maladies, sometimes providing insight into diseases etiology. For example, seasonal

changes in levels of allergens influence prevalence of allergies, while infectious diseases such as

flu follow a different seasonal pattern. Many recent examples of temporal clustering of disease

diagnoses suggest that discovery of seasonality is an important topic. Kawasaki disease, a

childhood vascular inflammation that can result in serious cardiac complication, has been recently

characterized by a distinct spatiotemporal distribution of cases (Burns et al. 2005; Rodó et al.

2011). In result, the still-uncharacterized pathogenic agent has been suggested to be a windborne

microbe. Anecdotal reports of seasonal and weather related patterns of disease incidence have

motivated studies on seasonality of heart failure (Gallerani et al. 2011), depression and anxiety

(Winthorst et al. 2011), varicose vein ulcers (Simka 2010), urinary tract infection (Anderson

1983; Falagas et al. 2009), and even cancer (Lambe, Blomqvist, and Bellocco 2003). While some

of these works searched for seasonality using purpose-driven surveys, Upshur (Upshur et al.

2005) used coded administrative data derived from a large EHR system to investigate whether

seasonal peaks in incidence were a common feature in a limited set of the most frequent

diagnoses. Some of these findings emphasize the behavioral causes of seasonal changes in

hospital visits, underlining the importance of attributing the likely biological versus sociological

causes of the patterns.

However, no systematic method has been developed to detect these seasonal patterns in an

unbiased broad scale. While some studies have searched for temporal patterns in disease

diagnosis, these works have been limited in the scope of the diseases examined and in the ability

to distinguish multiple types of novel seasonal patterns. The extensive longitudinal data on

diagnoses in the EHR is a unique source for finding trends in incidence of disease. However,

Page 108: Rachel D. Melamed

97

despite the promise of this data, and the potentially strong statistical power of these large patient

cohorts, inherent biases in these data obscure identification of seasonal trends. The most

computationally tractable component of the EHR is the ICD-9 code. These coded diagnoses are

primarily entered into the record in order to enable insurance billing, and entry is manual. Thus,

they are incomplete and the patterns of ICD-9 code entry may suffer a number of biases.

However, studies have assessed the ability of the ICD-9 to recover patients a wide range of

diseases, showing that they have a strong predictive value for diseases including skin infection

(Levine et al. 2013), urinary tract infection (Tieder et al. 2011), acute myocardial infarction

(Coloma et al. 2013), and chronic obstructive pulmonary disease (Stein et al. 2012).

Additionally, previous studies have examined the distribution of ICD-9 code entries over time in

order to learn characteristics unique to that disease. Temporal patterns in ICD-9 codes have been

used to pinpoint the influences in increased burdens to emergency units(Tang et al. 2010), and to

discover patterns in outcomes in high-risk surgeries(Finks, Osborne, and Birkmeyer 2011).

Members of our group have compared the well-characterized annual seasonal patterns in

influenza diagnoses against the influenza pandemic of 2009. The novel strain of the flu was found

to be associated with an unusual temporal distribution of influenza diagnoses(Khiabanian et al.

2010). The winter peak occurrence of viral in infections is well known; in contrast, bacteria may

cause more infections in warmer months(Perencevich et al. 2008). Thus, the EHR may contain

signals of seasonal incidence of disease, possibly implicating pathogens or other risk factors

influencing hospital admissions.

 Another advantage of seasonality as a research question is that this repeating pattern is less

influenced by the many biases inherent in ICD-9 codings. However, as described below, multiple

factors confound identification of periodicity. In the New York-Presbyterian EHR system, an

Page 109: Rachel D. Melamed

98

evident increase in diagnosis rates is obvious over the span of years observed. Additionally, the

total number of hospitalizations is itself influenced by season. After observing these

characteristics, we were motivated to propose a method (Lomb-Scargle periodograms in

detrended data, LSP-detrend) to correct for biases and robustly identify periodic temporal

patterns. To look for these patterns, we use Lomb–Scargle periodograms, a least-squares method

to detect periodic signal. LSP-detrend sensitively uncovers periodic temporal patterns after

applying these corrections to the data, and it assigns significance to the patterns. Subsequently,

we perform the first comprehensive survey of seasonality in hospital diagnoses, as reflected in

ICD-9 code incidence. We apply LSP-detrend to a compilation of records from 1.5 million

patients, comprising many million ICD-9 code entries. In result, we quantify the seasonal trend in

the 2,805 most common diagnosis codes coded over 12 years in the New York-Presbyterian

system.

Of these disorders, about 10% are identified as seasonal by LSP-detrend, including many known

phenomena. Performing a literature review on the resulting seasonal discoveries, we find that

many others of these confirm reported or well-established patterns, including some relatively rare

diseases. For example, we recover the seasonal winter increase in Kawasaki disease that has been

reported in other USA locations. One interesting novel finding is a bi-annual increase in acute

exacerbations of myasthenia gravis (ICD-9 code 358.01), with peak incidence in late winter and

late summer. This discovery has possible significance for this disease: acute exacerbation is a

serious, possibly life-threatening, complication of myasthenia gravis. Thus, we searched the EHR

for clues as to the cause of this seasonal pattern, using ADAMS(Holmes et al. 2011) to identify

diagnoses that are comorbid with the exacerbations in myasthenia gravis. We dissect the causes

of this seasonal incidence, proposing that factors predisposing patients to this event vary through

the year. Although EHR data, and ICD-9 coded records in particular, were not created with the

Page 110: Rachel D. Melamed

99

intention of aggregated use for research, these data can in fact be mined for periodic patterns in

incidence of disease, if confounders are properly removed. This work points to the potential of

the EHR as a source for unbiased pattern discovery, with implications for understanding human

disease.

5.2 Methods

5.2.1 Quantifying incidence of diagnoses We start from a de-identified data set from the New York Presbyterian EHR system, previously

used in(Holmes et al. 2011), that includes a patient number, date of visit, and diagnosis code.

After reviewing the total number of cases over the entire period recorded in the EHR, we restrict

our analysis to hospital visits happening from 1996, as the number of records drops significantly

before this year, until 2009, the last year available. From this data set, we extract all diagnoses

with more than 500 unique cases across that time period. Finally, the number of unique patients

presenting with the diagnosis per month comprises the input to the LSP-detrend analyses.

5.2.2 Correcting for confounding trends First, we examine the total trend of hospital visits by summing the number of cases of each

diagnosis together for each month. As shown in Figure 5-1A, and described in greater detail in

5.3.1, the number of cases increases steadily over time. These larger changes obscure the smaller

scale periodic pattern: periodograms measure the change from the mean signal as a function of

time. Before trend removal, no seasonal pattern is detected, but after the trend is removed, a

seasonal pattern in aggregate hospitalizations is evident (Figure 5-1B). Thus, the first step of

LSP-Detrend creates a smoothed version of this large scale pattern, representing the overall trend.

Subtracting the large scale trend from the data results in a “flattened” version of the diagnosis

data, with no large scale trend. The trends differ widely per diagnosis. For each code

individually, we calculate the smoothed trend at every month using the kernel density estimation

implementation from MATLAB which also estimates the appropriate bandwidth(Bowman and

Azzalini 1997). We remove the months two kernel bandwidths from the beginning and end of the

Page 111: Rachel D. Melamed

100

entire time period, as they cannot have reliable density estimates. Then, we subtract out the

smoothed estimate from the observations in order to create an incidence data set with no overall

trend.

The second step of LSP-Detrend removes the seasonal hospital visit trend, which is also

described in 5.3.1. This step makes use of the summed number of total hospital visits, once its

large scale trend is removed. For each diagnosis code, we mean scale this total hospital load to

match the mean number of cases of the diagnosis. This provides an estimate of number of

diagnoses of a disease per month that would occur if that diagnosis was always a fixed proportion

of the total hospital load. We subtract this monthly estimate of the seasonal hospital trend from

the “flattened” data to remove this overall hospital trend.

5.2.3 Evaluating periodicity The method that we chose for assessing periodic signal is the Lomb-Scargle periodogram. This

method was first developed for assessment of periodicity when temporal observations are

unevenly spaced(Lomb 1976; Scargle 1982). The computed periodogram evaluates the predictive

power of each tested frequency. Their work showed that the null distribution of the periodogram

for a frequency has follows an exponential, enabling assessment of statistical significance for a

given power (Scargle 1982). Using the corrected data described in the previous section, we apply

a MATLAB implementation of the Lomb-Scargle method(Press 1992). We discard any

significance assigned to periodic signals of with a more than 1.5 year period: these longer periods

are less interpretable and as the detrended data only spans a 10 year period, these signals are less

well supported by the data. We test 2,805 diseases for periodic patterns of incidence, and then we

use the Benjamini-Hochberg procedure (also implemented in MATLAB). We find that a Lomb-

Scargle p-value of < .01 has an expected false discovery rate of less than 15%.

Page 112: Rachel D. Melamed

101

5.2.4 Comorbidity analysis One interesting novel result is seasonality in acute exacerbations of myasthenia gravis, as

described in 5.3.4. We wish to characterize factors that might influence this seasonal pattern, by

looking for comorbid events in the EHR. In previous work from our group, my colleagues

developed an algorithm, ADAMS, for identifying comorbid disorders that are specifically

associated with a given query disease(Holmes et al. 2011). This method identifies diseases that

strongly co-occur with the query disease by comparing the levels of co-occurrence against a given

control disease. Finally, the method uses a bootstrap to estimate the false discovery rate.

For this application of ADAMS, we restrict co-occurring diseases to those diagnosed within 60

days before the acute exacerbation event, with a goal of capturing factors likely to have

immediate influence on, or to closely reflect, a patient’s state in the lead-up to this complication.

As ADAMS relies on the idea of comparing comorbidities against control diseases, we select

control diagnoses that capture aspects of these patients. Thus, we select controls with no likely

direct link to myasthenia gravis, but that occur in patient groups of similar age and additionally

are frequently diagnosed in this data set. The first control group is patients with influenza (code

487.1), as it is a very common disease that strikes a wide range of age groups in the winter. An

additional control group is patients with hip joint pain (719.45): this condition strikes patients

with a similar age distribution as myasthenia gravis, and, like the exacerbations of myasthenia

gravis, encounters of hip joint pain increase in the summer. The intersection of ADAMS

findings as found using each control provides our results, when conditions directly associated

with acute exacerbation are removed. The findings are discussed in 5.3.5, and listed in

Supplementary Table 5.

Page 113: Rachel D. Melamed

102

5.3 Results

5.3.1 LSP-detrend: finding periodic signal The EHR system at New York-Presbyterian hospital has been in place for three decades, and it

contains health information for 1.5 million patients, including both free text and coded entries for

diagnosis (ICD-9), procedures, prescription orders, and lab results. We select the time period

1997 to 2009 because the number of entries before 1997 falls off sharply in comparison. As very

rare diseases will not have enough data per month to infer a seasonal pattern, we select only

diseases with at least 500 cases over the period considered. The data set contains 2,805 diagnoses

with more than 500 cases and we obtain diagnosis date and patient identifier for each instance of

the diagnosis. The input to our method is then a count of the number of unique patients

diagnosed every month.

 

Two confounding trends emerge when we consider all diagnoses in aggregate. Identifying and

removing these trends, as described in 5.2.2, is a major step in identifying periodic signal. First, it

is clear that the number of patients visiting the hospital for any reason steadily increases over time

(Figure 5-1A). We remove this trend for each code by subtracting out a smoothed version of the

incidence information, a procedure we call de-trending. Upon removing this trend in the

aggregate diagnosis data, we are able to identify with strong confidence a seasonal increase in the

number of hospital visits in the spring and in the fall(Figure 5-1B). This hospital visit trend is

reflected in the monthly frequencies of the most common diagnoses: the more common a

diagnosis is, the more its monthly incidence reflects this overall trend (Figure 5-1C). The most

common diseases in the hospital include many chronic diseases, such as Unspecified Essential

Hypertension (401.9), Obesity unspecified (278.00), and Osteoporosis unspecified (733.00). The

high prevalence of chronic diseases among the diseases with a spring and fall increase provides a

clue as to the meaning of this trend. These diseases are unlikely to be the primary cause of most

Page 114: Rachel D. Melamed

103

seasonal hospital visits. A more likely explanation is that hospital visits increase seasonally for a

number of reasons and these common diseases simply represent a fixed proportion of the overall

population. Thus, in a procedure that we term de-totaling, also described in 5.2.2, we remove the

total population trend.

Figure  5-­‐1:  Identifying  confounding  factors  in  temporal  diagnosis    

A.    Aggregated  number  of  diagnoses  from  1997  to  2009  (blue)  show  a  strong  increasing  trend  over  time,  modeled  by  the  red  line.    When  this  trend  is  subtracted  out,  the  remaining  signal  (magenta)  shows  no  overall  trend  but  a  seasonal  trend.  B.  De-­‐trended  total  diagnoses  display  6  month  periodicity,  as  shown  by  the  periodogram.    When  the  years  are  plotted  on  top  of  each  other  (each  colored  line  represents  a  year,  with  the  bold  black  line  the  average),  the  spring  and  fall  show  consistent  peaks  in  diagnosis  each  year.  C.  For  each  diagnosis,  the  number  of  cases  is  compared  with  the  seasonal  pattern  in  incidence.    The  most  frequently  occurring  diagnoses  show  the  most  correlation  with  the  overall  spring-­‐fall  peak  incidence;  the  overall  trend  causes  false  detection  of  periodic  signal.  Correlation  between  number  of  cases  (x-­‐axis)  and  correspondence  with  the  overall  spring-­‐fall  trend  (y-­‐axis)  is  0.41.  

0 5 10 15 200

5

10

15

20Period 0.5 yrswith FAP of 9.0172e−07

0 5 10 15 200

5

10

15x 1010

Period = 0.5

1996 1998 2000 2002 2004 2006 2008 20100.6

0.8

1

1.2

1.4

1.6

1.8

2x 105

0 5 10 15 200

5

10

15

20Period 0.5 yrswith FAP of 9.0172e−07

0 5 10 15 200

5

10

15x 1010

Period = 0.5

1996 1998 2000 2002 2004 2006 2008 20100.6

0.8

1

1.2

1.4

1.6

1.8

2x 105

Period (years)

A

B

Total diagnoses Trend to remove De-trended total

C

pow

er

Number of cases (x 10,000)

Cor

r. w

ith to

tal t

rend

Page 115: Rachel D. Melamed

104

The final step of LSP-Detrend assesses the adjusted data for periodic signal. We use Lomb-

Scargle periodograms, which use the time series of monthly rate of diagnosis as the input. For a

range of possible periods, the power of that period, and an associated statistical significance, is

calculated. We find the period of greatest power for the uncorrected data, the de-trended data,

and the de-trended and de-totaled data.

5.3.2 Major types of periodic signal and known seasonal disease Of the 2,805 codes in our study, 284 have a significant periodic signal that is likely to represent

seasonal peaks in incidence. Performing hierarchical clustering of the monthly occurrence of

each disease, we look for groups of conditions with similar period and phase of incidence (Figure

5-2). The clustering shows that two main groups comprise most of the diagnoses.

Figure  5-­‐2:  Pre-­‐processed  and  row-­‐normalized  monthly  incidence  for  227  codes  with  periodic  signal.  

 Each  row  is  a  disease,  and  each  column  a  month  over  10  years.  Thus,  boxes  in  a  row  represent  incidence  of  that  disease  for  each  month,  with  red  signifying  elevated  incidence  and  green  decreased  incidence.    Two  main  clusters  stand  out:  diseases  that  occur  in  the  summer  (top),  and  those  that  occur  in  winter.  

Month (ordered by time over 10 years)

Dis

ease

(clu

ster

ed)

Sum

mer peak

Below average

Above average

Winter peak

Page 116: Rachel D. Melamed

105

The two groups split between events that occur in winter (including viral infections and

respiratory infections), and those diagnoses that occur in summer (mostly fractures and wounds).

Based on the groupings of codes, seasonal influences appear to arise from a number of sources.

 

Many of the patterns appear to be associated with seasonally influenced behavior changes.

However, the majority of these have not been previously reported in the literature. This includes

the predominance of accidents in the summer. As well, rashes and skin infections like impetigo

are likely linked to more skin exposure in the summer, and the same factor seems likely to

explain the increased diagnosis of the chest bone malformity pectus excavatum (code 754.81).

Seasonal behavior change also drives the pattern in diagnoses pertaining to child psychiatric

disorders, including attention deficit disorder and adjustment disorders, which dip sharply during

the summer school break. A well known annual pattern, the increase in births in the summer, has

also been suggested to be most attributable to behavior, though other factors may play a role

(Bobak 2001).

Although seasonal changes in behavior explains many of the temporal patterns in diseases rates

throughout a year, environmental risk factors clearly vary as well, including allergens, ultraviolet

light, and the virulence of pathogens. It is well known that some allergies, influenza, pneumonia,

scarlet fever, and complications from these disease have clear peaks in incidence. All of these

effects were captured in our data and by LSP-detrend, showing that ICD-9 codes do reflect these

patterns, and that our method is sensitive to such signal. The next section focuses on the findings

that appear most novel, interesting, and interpretable.

5.3.3 Confirmation of recent reports of seasonal effects  While no previous method has performed a systematic search for seasonal trends in disease

incidence, some previous studies have assessed seasonality of individual diseases. These studies

Page 117: Rachel D. Melamed

106

are usually inspired by anecdotal observation of a seasonal association of a disease, and these

reports use a variety of methods to scientifically assess these hypotheses. such as questionnaires,

mining surveillance databases, examining lab results, or performing patient chart review for a

pre-selected set of patients. LSP-Detrend, on the other hand, does not require a pre-defined

hypothesis, but instead it is able to rank the seasonality for all diagnoses in the hospital, using

existing hospital data. While the previous studies have required extensive data collection, we take

advantage of an already rich data source, and we show that LSP-Detrend, and the ICD-9 data, is

able to reproduce findings reported elsewhere. The diseases that LSP-Detrend assigns strong

signal include a number of previous reports of periodic disease incidence. We compare our

findings to these studies below.

First, neuropsychiatric diseases provide a particularly interesting subset, as influences in their

occurrence are controversial. Some literature supports seasonal changes in occurrence of anxiety

and depression (Winthorst et al. 2011). Taking a mechanistic approach, other groups have

documented seasonality of key neurotransmitters involved in mood (Lambert et al. 2002;

Molendijk et al. 2012). Our analysis uncovers a strong winter and early spring increase in

obsessive-compulsive disorder (300.3), dysthymic disorder (300.4), shown in Figure 5-3A, and

other depressive disorders (311). It is difficult to attribute trends in these complex disorders to

behavioral versus environmental influences. Thus, we find an interesting contrast in other

psychiatric disorders, such as dependent personality disorder and social phobia. These have no

seasonal pattern, suggesting that different factors influence these diseases.

Page 118: Rachel D. Melamed

107

Figure  5-­‐3:  Selected  diseases  with  periodic  signal.    

For  four  of  the  diseases  discussed,  the  monthly  incidence  for  all  years  is  plotted  together  in  order  to  view  consistencies   in   the   seasonal   trend   across   different   years.     Each   colored   line   again   represents   a   year,  with  the  bold  black  line  the  average  across  the  years.  

Other periodically increasing diseases are likely linked to seasonally increased environmental

risk. Recently, reports have asserted that bacterial infections are more frequent in warmer weather

(Perencevich et al. 2008), that bacterial bloodstream infections peak in summer (Eber et al. 2011),

and that there is a strong seasonal significant effect on bacteria virulence (Frankel et al. 2012).

Our data strongly support the hypothesis that bacterial infections are higher in the summer. We

detect strong summer periodic signal in urinary tract infection (code 599.0), shown in Figure

5-3B, and its complications of pyelonephritis (590.10 and 590.80), and hematuria (599.7). This

corroborates results from (Anderson 1983; Falagas et al. 2009). We also detect increased rates of

cellulitis and abscess in the summer months. Other groups have found increased incidence of soft

tissue infections in the summer (X. Wang et al. 2013). Finally, there is a strong late summer peak

incidence of vascular device inflammation and infection (code 996.62), which may be due to the

same influences.

0 2 4 6 8 10 12 14 16 18 20 220

10

20

30

40

Period 0.97778 yrswith FAP of 8.8833e−09

0 2 4 6 8 10 12 14 16 18 20 220

0.5

1

1.5

2 x 107

Period = 1

1998 2000 2002 2004 2006 2008600800

1000120014001600

de−trend, de−totaled 599.0 Urinary tract infection site not specified cc = −0.13

J F M A M J J A S O N D J900

1000

1100

1200

1300599.0 Urinary tract infection site not specified

0 5 10 15 200

5

10

15

20

Period 1.0427 yrswith FAP of 0.00040641

0 5 10 15 200

0.5

1

1.5

2 x 106

Period = 1.0167

1998 2000 2002 2004 2006 2008100

200

300

400

500

de−trend, de−totaled 300.4 Dysthymic disorder cc = 0.12

J F M A M J J A S O N D J100

200

300

400300.4 Dysthymic disorder

0 1 2 3 4 5 6 70

5

10Period 0.50575 yrswith FAP of 0.0094043

0 1 2 3 4 5 6 70

500

1000

1500

2000 Period = 0.52381

2004 2005 2006 2007 2008 2009

5

10

15

de−trend, de−totaled 358.01 Myasthenia gravis with (acute) exacerbation cc = −0.08

J F M A M J J A S O N D J0

5

10

15358.01 Myasthenia gravis with (acute) exacerbation

0 2 4 6 8 10 12 14 160

2

4

6

8

10Period 0.9899 yrswith FAP of 0.0093922

0 2 4 6 8 10 12 14 160

2000

4000

6000

8000

10000

Period = 1.0104

1998 2000 2002 2004 2006 2008

5

10

15

de−trend, de−totaled 446.1 Acute febrile mucocutaneous lymph node syndrome (mcls) cc = −0.08

J F M A M J J A S O N D J0

5

10

15

20446.1 Acute febrile mucocutaneous lymph node syndrome (mcls)A

B

C

D

Page 119: Rachel D. Melamed

108

Finally, as one inspiration for this work is to discover novel pathogenic effects contributing to

disease incidence, we are particularly interested in the identified annual winter peak in Acute

febrile mucocutaneous lymph node syndrome (mcls) (code 446.1), also known as Kawasaki

disease. The winter peak in incidence (Figure 5-3C) is strong and the finding is consistent with

previous USA reports(Rodó et al. 2011). Although this disease is rather infrequent in our cohort,

LSP-detrend confidently identifies the pattern.

Thus, the results contribute knowledge of a range of human diseases, and many of our findings

are buttressed by previous reports investigating specific hypotheses about disease incidence.

5.3.4 Novel findings: acute exacerbations of myasthenia gravis One of our findings stands out for further analysis. Myasthenia gravis with acute exacerbation,

code 358.01, was particularly interesting because it is a well-defined diagnosis, the condition is

acute, requiring immediate attention, and the seasonal incidence is previously entirely unreported.

Although this is a rare condition, the seasonal trend is strongly visible in Figure 5-3D, with peak

incidences in late winter and in late summer.

Myasthenia gravis is an autoimmune disease characterized by presence of antibodies targeting

elements of the nerve to muscle junction. The result is a decay of this neuromuscular junction,

resulting in blocked neural signals to the muscle and subsequent muscle weakness (Querol and

Illa 2013). Subtypes include patients with antibodies targeting the acetylcholine receptor, and

against the muscle specific kinase receptor, with variable phenotype, including treatment

response, depending on the category of autoimmune antibody. Patients, usually older middle-

aged people or young women, often primarily present with eyelid weakness (ptosis) or difficulty

swallowing (dysphagia) or other signs of weakness. Suggested underlying causes include

abnormalities of the thymus, such as thymoma, certain drugs, as well as a genetic component.

Page 120: Rachel D. Melamed

109

Although immunosuppressive drugs or thymectomy can mitigate the weakness and allow a high

quality of life for most subtypes of patients, the condition is incurable and is characterized by

occasional acute episodes. Any infection can cause increased immune activity which can worsen

an autoimmune condition, and the role of viral infections in particular have been investigated

(Cufi et al. 2013). Other risk factors include stress, many drugs, and temperature, which directly

reduces the efficiency of the neuromuscular junction. In acute exacerbation of myasthenia gravis,

a patient experiences greater weakness, sometimes to the point where normal respiratory activity

is threatened, resulting in respiratory failure. Thus, patients with this diagnosis have a high

likelihood of seeking prompt treatment. The chronic condition, with a different administrative

code, shows no seasonal pattern.

5.3.5 Dissecting causes of seasonality of acute exacerbations by finding comorbid diseases

As this seasonal pattern has no obvious cause, we use the EHR data to search for common factors

among patients with exacerbation. Concurrent diagnoses, prescriptions, or procedures could

provide insight into cause of the seasonal pattern. Described previously (Holmes et al. 2011),

ADAMS, Application for Discovering Disease Associations using Multiple Sources, has been

used with the same clinical data set to find comorbid diagnoses for rare diseases. We use the

method to search for comorbid events specific to the months preceding the exacerbation, as

described in 5.2.4, and the results are shown in Supplementary Table 5. When conditions that can

be assumed to be a direct result of the myasthenia gravis are disregarded, the most associated

diseases are urinary tract infection (code 599.0), carpal tunnel syndrome (code 354.0),

unspecified essential hypertension (code 401.9) and esophageal reflux (code 530.81). We find it

particularly interesting that both UTI and carpal tunnel are seasonally linked, although the

explanation for the pattern in carpal tunnel is unclear. It is possible that treatment for a seasonally

linked disease exacerbates the myasthenia gravis, as this condition is worsened by many

commonly used drugs.

Page 121: Rachel D. Melamed

110

5.3.6 Comparison between hospital systems Because ICD-9 is a standard, we finally sought to reproduce our seasonal findings in the Stanford

EHR system. The Stanford data covers a later time period than the Columbia data did, comes

from a geographic area with a more moderate climate, and contains a patient group that is not

likely to have strong overlap with the Columbia data set. We find that this system has a much

weaker seasonal pattern of overall hospitalizations. While the seasonal pattern is correlated to

that observed in Columbia,, the variance between years was much higher (Figure 5-4).

 Figure  5-­‐4:  Overall  seasonality  of  hospitalization  in  Columbia  and  Stanford  

On  the  left,  the  data  across  years  is  plotted,  showing  the  de-­‐trended  total  hospitalizations  in  pink.    On  the  right,  each  year  is  plotted  over  the  others,  with  the  mean  across  the  years  plotted  as  the  thicker  black  line.  

Approximately 25% of the seasonal ICD-9 codes from the Columbia system have similar patterns

of significant seasonality in the Stanford system, and an additional number of similar diseases

have similar seasonal patterns between the two hospitals. Of note, clear coding differences exist

between the hospitals, confounding a full comparison of the patterns. For example, where the

Columbia system has a seasonal increase in urinary tract infection, the Stanford system shows a

0 5 10 15 200

5

10

15

20Period 0.5 yrswith FAP of 9.0172e−07

0 5 10 15 200

5

10

15x 1010

Period = 0.5

1996 1998 2000 2002 2004 2006 2008 20100.6

0.8

1

1.2

1.4

1.6

1.8

2x 105

Total&diagnoses&Trend&to&remove&De1trended&total&

COLUMBIA&

STANFORD&

Page 122: Rachel D. Melamed

111

similar increase in acute cystitis. A number of diseases, including myasthenia gravis with acute

exacerbation, have too few patients in the Stanford system for the LSP-Detrend analysis. Other

similarities between the two sets of results include: a springtime increase in carpal tunnel and

ulnar nerve lesion in both sets; a summer increase in accidents and birth related events; and an

increase in complications of procedures in the summer. We conclude that there is some

reproducibility, but that it is difficult to confidently identify patterns in rare diseases, and that

ICD-9 can limit identification of patient cohorts.

 5.4 Discussion

Applying periodograms to administrative code data from EHR, we are able to identify a number

of periodic patterns in an unbiased fashion. Importantly, we demonstrate that examining the data

for confounding trends is an essential step for this research question. When the proper

corrections are made, the results confirm that LSP-detrend is sensitive to expected seasonal

variation, and the method also provides support for recent findings of seasonal distributions of

disease. Most significantly among these is a pervasive pattern of increased incidence of bacterial

infections in the warmer season, including urinary tract infection, cellulitis and abscess, as well as

infection and inflammation of vascular implant. Although some community-acquired bacterial

infections in fact are more frequent in the wintertime, it is possible that a distinct subset of

bacteria, relying less on community transmission, show more virulence in the warmer weather.

The finding that myasthenia gravis exacerbation has a periodic increase in incidence will be of

interest both to clinicians providing care to the patients as well as to immunologists seeking to

understand the conditions in which the autoimmune disease is worsened. The results of the

comorbidity analysis show that urinary tract infection in particular, as a strong covariate with the

condition and as a seasonally linked disease, may have a role to play in exacerbation. It is known

Page 123: Rachel D. Melamed

112

that some antibiotics, including those for treatment of urinary tract infection, can worsen the

condition. However, as no single correlating factor explains the seasonal pattern clearly, this may

be an interesting avenue for further research. There may be an underlying infectious agent

causing the immune system flare.

Although ICD-9 codes are primarily recorded for administrative purposes, they have a number of

advantages for use in research. Coded data provides a clear categorization of patients and thus

this data is a suitable starting point for well-developed computational methods, as compared to

other types of EHR information that require inference of disease state. The ICD-9 represent a

wide variety of disease, and, importantly, they are an international standard. Thus, a method

developed with our system in New York can easily be applied to the myriad other large EHR

datasets across the world. In the analysis performed at Stanford, many highly reproducible

findings with clinical significance were uncovered. This implies that despite the limitations of

ICD-9 codes, code-based EHR studies offer a promising avenue of research.

EHRs are an increasingly rich source of information. In the future, projects such as the eMERGE

network promise the integration of this phenotypic data with genotypic information (Gottesman et

al. 2013). With the advent of these resources, patients who display increased encounters for a

disease could be interrogated for genotypic markers, allowing us to find new mechanisms of

disease, as has been previously investigated in psychiatric disorders. The appropriate methods to

analyze this complex source of information in an unbiased fashion holds great promise for human

disease.

Page 124: Rachel D. Melamed

113

6 CONCLUSION

The approaches I have developed in my PhD studies consider that cancer is the result of a

combination of aberrant processes, and signatures of these underlying processes can be found

within large collections of genomics and clinical data. Patterns in incidence of cancer are the best

source of existing data on cancer biology, yet most methods that mine these data sources have

only considered certain types of patterns as interesting. These include signals such as: recurrent

selection of mutation(Lawrence et al. 2013; Mermel et al. 2011), mutual exclusivity of

mutation(Ciriello et al. 2011; Vandin, Upfal, and Raphael 2012), regulatory relationships(Akavia

et al. 2010; Margolin et al. 2006); and projection of genomic data onto annotated gene functional

relationships (Sedgewick et al. 2013; Mark D M Leiserson et al. 2014; Hofree et al. 2013).

A unifying theme of my dissertation is finding new ways to integrate the overwhelming variety

and quantity of cancer data available. My work has been aimed at expanding the horizon of

meaningful patterns in cancer data. In my projects using the total correlation score to find sets of

genetic alterations with a related pattern of occurrence, I develop a highly general approach to

looking for non-random modules of mutated genes in this data. First, in 3.1, I use total

correlation to discover modules in brain neoplasms, an approach that I show can indicate

underlying distinctions in the biology of tumor subtypes. Then, in 3.2, I expand on this premise

by developing a method that does not use recurrence at all, but instead can identify cancer drivers

only by their strong related pattern of alteration with other genes. While cancer is caused by

heterogeneous patterns of somatic alterations, Mendelian diseases stand in sharp contrast: these

are caused by quite homogeneous and highly penetrant germline variants. But germline mutations

that predispose patients to develop particular cancers may, like somatic mutations, represent

cellular processes contributing to cancer growth. In chapter 4 I show that comorbidity uncovered

Page 125: Rachel D. Melamed

114

from clinical records can be related to processes shared between Mendelian variants and somatic

mutations in the comorbid cancer. This finding adds another dimension to TCGA data.

The novel results I have described in this dissertation are also unified by their potential to not

only identify key drivers among mutated genes, but to provide insight into the roles of candidate

drivers in cancer progression. The results all place genes in the context of related alterations. The

total correlation projects mine this large high-quality genomics data for combinatorial patterns of

mutations. For example, using GAMToC, I discover a strong connection between TP53 mutations

and deletions in the BRSK2 locus on chromosome 11, co-occurring mutations that have not been

previously highlighted in literature. Adding Mendelian disease comorbidity allows us to compare

the action of disease genes in the context of multiple diseases. In the results from comorbidity

and genetic similarity analysis, one interesting gene is PTK6, a recurrently amplified gene in the

melanoma cohort that may influence mesenchymal transition in epithelial cells. Among the

recurrently altered melanoma genes, this gene stands out for its connection to genes associated

with the melanoma-comorbid Mendelian disease epidermolysis bullosa. The epidermolysis

bullosa variants impact dermal adhesion and maintenance of basement membrane. The specific

phenotype of the Mendelian disease therefore further suggested that in melanoma, PTK6 and the

epidermolysis bullosa genes might play a role in invasion and metastasis. This hypothesis was

supported in subsequent analysis.

My dissertation work has focused mainly on genetic mutations, specifically copy number

alterations and coding sequence mutations. However, many other dimensions are available in

cancer genomics data: gene and microRNA expression, methylation, fusion patterns, reverse-

phase protein array expression. I have already shown that gene expression patterns of sets of

Mendelian disease genes demonstrate informative signal in their comorbid cancers. This approach

Page 126: Rachel D. Melamed

115

used gene expression, and clinical information, as a validation of our candidate Mendelian-related

cancer drivers. But, an important improvement on the Mendelian disease project could use this

data to predict more candidates. It would also be interesting to apply this idea in the context of

the GAMToC projects: sets of co-expressed genes could be identified, and then the combination

of sets of gene expression could be assessed using our entropy-based score.

Results from both of my main dissertation projects have provoked further questions that I would

like to pursue in my future work. Related genetic mutation events, such as those detected by

GAMToC, are signifiers of convergence in the evolution of cancers across patients. To return to

the example of the cell cycle genes mutated in glioblastoma, different subtypes of this disease

appear to have specific associations with either CDK4, RB1, or CDKN2A alterations. These genes

are all tightly functionally connected, and they have a mutually exclusive pattern of occurrence

across the glioblastoma data. While some studies have claimed that this pattern is evidence that

the mutations represent alternative equal effects, our result places each of these mutually

exclusive alterations in context of other associated mutations. Far from being redundant events,

these mutations appear to have specific ramifications affecting cancer progression in different

mutational contexts. This is known to be true for changes to the CDKN2A locus, which also

influence the TP53 pathway. This is a hint that subtypes of cancer represent combinations of

cancer sub-programs, perturbing specific biological processes. Compelling the cellular

machinery into a continuous process of division is a shared characteristic that most tumors

evolve. But how this function is acquired varies widely. Cancers could acquire this function as a

result of germline variants, in the case of Mendelian disease patients with increased cancer risk.

In the case of cancer subtypes, subtype-specific genetic alterations could endow tumors with the

needed trait. But what is clear is that the cell cycle trait is “conserved” across tumor cohorts.

Page 127: Rachel D. Melamed

116

Most current approaches (Hoadley et al. 2014; Mo et al. 2013; Akbani et al. 2014) searching for

signs of convergent evolution across tumors look for hidden classes of tumors. A common

approach, achieved through widely varying means, is to cluster tumors regardless of tissue type.

In contrast, very few methods, such as PARADIGM (Sedgewick et al. 2013), quantify the

processes in each tumor individually. As each tumor evolves through random mutation followed

by selection, treating tumors as members of unified subclasses is a limited approach that can miss

the interesting exceptions to this rule. But because tumors clearly have much in common,

treating each cancer as unrelated to the others also neglects a large source of information.

Therefore, an approach that I hope to pursue would involve treating individual tumors as a

combination of events that are similar to those present across multiple tumors. In this manner, we

can discover new shared events driving cancer development, and we can understand how these

events impact an individual tumor, with potential therapeutic implications. For example, in

melanoma the “Mendelian code” of comorbidities indicates pathways such as melanocyte

differentiation and dermal adhesion, reflecting the cell of origin and its surrounding environment.

Growing brain neoplasms face a different set of challenges. As the brain cells they arise from

have low inherent replicative potential, alterations related to telomerase activity may be essential

to these cancers. Mutations impacting telomere maintenance in brain neoplasms are widespread,

but occur in diverse, and usually mutually exclusive, fashion including ATRX coding mutations,

TERT promoter mutations, or the alternative lengthening of telomeres mechanism(Remke et al.

2013). We can harness patterns across tumor cohorts to better understand the mechanisms

underlying development of each individual tumor. In this way, the results from my dissertation

work can be connected to the goals of precision medicine.

More broadly, patterns in cancer data will only become of increasing importance as the amount

and type of cancer data grows. New sources of data such as the eMERGE consortium

Page 128: Rachel D. Melamed

117

(Gottesman et al. 2013) have enabled interrogation of the connections between genotypes

influencing multiple diseases, an approach that has been termed phenome-wide association

study(Denny et al. 2013). Eventually, we will have a wider array of such genotypic and

phenotypic information for cancer patients and for the population at large. We will soon be able

to integrate a cancer patient’s somatic mutations and clinical trajectory with possible germline

influences and environmental factors. At the other end of the spectrum, recent studies have

sequenced subclonal populations of tumors at the single cell level. Patterns in subclonal evolution

of tumors would be expected to follow many of the same principles as are found at the population

level. Convergence in this evolutionary process can be identified to find cancer drivers, much as

we did with the GAMToC projects. Additionally, combinations of co-occurring subclonal

populations can help us understand the mutation profile observed in bulk tumor data. Cancer

genomics, and methods to find patterns impacting tumor growth, has potential to decode the

complex series of events leading to cancer, telling us about the biology of tumors and thus the

treatments that will provide greatest impact.

Page 129: Rachel D. Melamed

118

           

7 Supplementary Tables Supplementary  Table  1:    Significantly  mutated  genes  in  the  melanoma  cohort,  and  their  mutations  across  the  tumors  

   

pat tumor(Var(Depthtumor(Pos((Depthnormal(Var(Depthnormal(Pos((Depthchr:pos3pos r/f genes codons AAspatient36 24 51 0 32 chr7:1404531363140453136A/T BRAF GTG2GAG V600Epatient073057 18 50 0 23 chr7:1404531363140453136A/T BRAF GTG2GAG V600Epatient073058 9 24 0 34 chr7:1404531363140453136A/T BRAF GTG2GAG V600Epatient073232 28 48 0 30 chr7:1404531363140453136A/T BRAF GTG2GAG V600Epatient103104 27 53 0 29 chr7:1404531363140453136A/T BRAF GTG2GAG V600Epatient103276 42 70 0 29 chr7:1404531363140453136A/T BRAF GTG2GAG V600Epatient145 12 46 0 90 chr2:1187328043118732804G/A CCDC93 GCT2GTT A237Vpatient16 10 35 0 55 chr2:1187159843118715984G/A CCDC93 TCC2TTC S321Fpatient073232 25 46 0 62 chr1:2068217303206821730G/A DYRK3 GGC2GAC,GGC2GACG396D,G376Dpatient103104 15 43 0 38 chr1:2068210603206821060C/T DYRK3 CCA1TCA,CCA1TCAP173S,P153Spatient073057 12 38 0 54 chr4:1532473453153247345C/T FBXW7 TGG2TAG,TGG2TAG,TGG2TAGW368*,W486*,W406*patient073058 8 25 0 47 chr4:1532473453153247345C/T FBXW7 TGG2TAG,TGG2TAG,TGG2TAGW368*,W486*,W406*patient16 17 41 0 73 chr12:784288137842881G/A GDF3 CAT1TAT H230Ypatient36 12 42 0 27 chr12:784311537843115C/T GDF3 GAG1AAG E152Kpatient16 21 51 1 140 chr12:14798222314798222G/A GUCY2C CCT1TCT P580Spatient36 13 28 0 35 chr12:14829863314829863C/T GUCY2C ATG3ATA M291Ipatient103276 14 37 0 26 chr12:14809526314809526G/A GUCY2C CGT1TGT R464Cpatient16 37 68 0 121 chr2:1029553453102955345C/T IL1RL1 CCT2CTT,CCT2CTTP37L,P37Lpatient103276 10 38 0 37 chr2:1029655423102965542G/A IL1RL1 GGA2GAA G374Epatient073057 7 23 0 30 chr5:35876230335876230G/A IL7R GGA2GAA G341Epatient073058 6 24 0 18 chr5:35876230335876230G/A IL7R GGA2GAA G341Epatient103104 16 47 0 32 chr17:51900681351900681C/T KIF2B TCC2TTC S96Fpatient103276 10 19 0 20 chr17:51900950351900950G/A KIF2B GAA1AAA E186Kpatient145 20 71 1 78 chr6:63990011363990011C/T LGSN CGA2CAA R482Qpatient16 20 41 0 151 chr6:63990305363990305C/A LGSN TGG2TTG W384Lpatient16 7 22 0 81 chr6:63991041363991041T/C LGSN AGA1GGA R139Gpatient16 44 97 0 188 chr4:1642720423164272042C/T NPY5R TCA2TTA S206Lpatient103104 29 83 1 90 chr4:1642719573164271957C/T NPY5R CAC1TAC H178Ypatient145 17 52 0 47 chr8:32621309332621309G/A NRG1 GAT1AAT,GAT1AAT,GAT1AATD443N,D435N,D438Npatient103104 26 82 0 62 chr8:32453481332453481G/A NRG1 CGA2CAA,CGA2CAA,CGA2CAA,CGA2CAA,CGA2CAAR294Q,R79Q,R79Q,R79Q,R79Qpatient36 7 21 0 13 chr11:479109034791090G/A OR51F1 CCT1TCT P20Spatient103104 11 20 0 46 chr11:479061634790616G/A OR51F1 CAC1TAC H178Ypatient145 24 66 0 85 chr7:1424585443142458544C/T PRSS1 TCA2TTA S60Lpatient16 42 167 0 260 chr7:1424584203142458420G/A PRSS1 GAT1AAT D19Npatient073058 26 95 0 98 chr9:33796691333796693GAG/3 PRSS3 del E(88388)3patient073232 16 74 6 100 chr9:337979283337979283/C PRSS3 AGG2+,AGG2+R158+,R101+patient145 24 52 1 75 chr13:32376340332376340C/T RXFP2 CCA2CTA P688Lpatient103276 7 32 0 33 chr13:32365959332365959C/T RXFP2 CGA1TGA R388*patient16 14 39 0 73 chr2:2188703218870C/T SH3YL1 GAA1AAA E228Kpatient073057 11 24 0 41 chr2:2310823231082C/T SH3YL1 GAA1AAA E119Kpatient073058 6 27 0 36 chr2:2310823231082C/T SH3YL1 GAA1AAA E119Kpatient36 17 40 0 23 chr19:52002863352002863T/G SIGLEC12 ACG1CCG T306Ppatient073057 7 15 0 18 chr19:51995082351995082C/T SIGLEC12 GGA2GAA G534Epatient16 12 24 0 48 chr6:13588577313588577C/T SIRT5 CGA1TGA,CGA1TGAR44*,R44*patient073058 39 104 0 149 chr6:13612078313612078A/G SIRT5 GAA2GGA E305Gpatient36 50 175 1 148 chr3:39432984339432984C/T SLC25A38 TCT2TTT S110Fpatient073232 30 39 0 105 chr3:39433013339433013C/T SLC25A38 CCC1TCC P120Spatient145 13 44 0 41 chr20:42694523342694523G/A TOX2 GGC1AGC,GGC1AGC,GGC1AGCG336S,G360S,G378Spatient145 13 44 0 40 chr20:42694524342694524G/A TOX2 GGC2GAC,GGC2GAC,GGC2GACG336D,G360D,G378Dpatient103104 13 27 0 33 chr20:42635250342635250C/T TOX2 CTC1TTC,CTC1TTC,CTC1TTCL35F,L86F,L77Fpatient103276 16 31 0 16 chr20:42682943342682943C/T TOX2 TCG2TTG,TCG2TTG,TCG2TTGS177L,S228L,S219Lpatient073057 14 45 0 48 chr2:2345454183234545418G/A UGT1A10 GAA1AAA E84Kpatient073058 11 45 0 41 chr2:2345454183234545418G/A UGT1A10 GAA1AAA E84Kpatient145 7 18 0 38 chr3:1672489563167248956T/A WDR49 AAT2ATT N370Ipatient16 11 27 0 38 chr3:1672457473167245747G/A WDR49 TCA2TTA S470L

Page 130: Rachel D. Melamed

119

Supplementary  Table  2:  Genes  significantly  less  frequently  mutated  in  the  nevus  cohort,  see  2.2.2.4

 

 

melF(279) nevF(32) p mutsigTHSD7B 151 0 1.09E=10 0.1663559XIRP2 173 2 4.46E=10 0.04012697USH2A 147 1 7.76E=09 0.04012697PTPRT 113 0 1.96E=07 0.09801548MYH1 99 0 2.09E=06 0.1379674DSP 90 0 8.81E=06 0.1761339TPTE 88 0 1.20E=05 0.1529721SYNE1 114 2 3.08E=05 0.1158377KCNB2 79 0 4.69E=05 0.1962425COL3A1 72 0 0.00013 0.00882006PCDH18 71 0 0.00015 0.03796533BCLAF1 62 0 0.00052875 0.0220468DSG3 62 0 0.00052875 0.03248277PDE4DIP 62 0 0.00052875 0.167958SNCAIP 61 0 0.00060636 0.1663559ROS1 77 1 0.00072576 0.1954481TCHHL1 54 0 0.00155643 0.01057947ACSM2B 53 0 0.0017768 0.03575059TP63 53 0 0.0017768 0.03657042ANO4 51 0 0.00231181 0.03657042NBPF1 44 0 0.00571202 0.01783554NRK 44 0 0.00571202 0.00548988THEMIS 43 0 0.00648653 0.09676519ARID2 42 0 0.00736234 1.24E=07STK31 42 0 0.00736234 0.1947338RUNX1T1 41 0 0.00835224 0.167958TP53 41 0 0.00835224 4.57E=13NF1 40 0 0.00947053 2.53E=10NBEAL1 39 0 0.01073327 0.1407236PDE1A 39 0 0.01073327 0.00039644ADAM30 38 0 0.01215843 0.1407236KEL 38 0 0.01215843 0.01783554SELP 37 0 0.01376616 0.06169981POTEG 36 0 0.01557899 0.169217SLC38A4 35 0 0.01762213 0.00190358MPP7 34 0 0.0199238 0.07358765NRAS 85 4 0.02196215 4.57E=13OR51S1 33 0 0.02251551 0.02791723CDKN2A 31 0 0.02871411 4.57E=13EPHA3 31 0 0.02871411 0.1407236MLL 43 1 0.03979048 0.1262187NFASC 27 0 0.04644532 0.1179532

Page 131: Rachel D. Melamed

120

Supplementary  Table  3:  Pairs  of  comorbid  and  genetically  similar  Mendelian  disease  and  cancer,  related  to  4.3.    Columns  described  below:    

gene_enriched:  The  corrected  significance  of  the  number  of  genes  in  common.              

geneIntersection:  This  shows  the  common  genes,  even  if  not  statistically  enriched          

pathway_correlation  and  pathway:  "pathway_correlation"  shows  the  Spearman  p-­‐value  for  the  pathways  correlation  for  the  pair  of  diseases,  after  correction  for  427  tests.  If  this  is  less  than  .1,  and  if  there  are  any  shared  pathways  (significantly  enriched  in  the  cancer  and  impacted  by  the  Mendelian  disease),  the  pathways  are  shown  in  the  "pathways"  column.    The  format  for  each  pathway  shared  is:  Mendelian_gene_1_in_pathway,  Mendelian_gene_2_in_pathway  -­‐>  Pathway_name  (Cancer_gene_1_in_pathway,  Cancer_gene_2_in_pathway);...          

coex_CG  and  coexpression:  "coex_CG"  shows  the  best  coexpression  score  for  the  pair,  corrected  across  all  coexpression  results.    If  this  corrected  values  is  less  than  .1,  all  cancer  genes  showing  coexpression  with  the  Mendelian  disease  genes  are  shown  in  the  "coexpression"  column.      Each  significant  cancer  gene,  is  displayed  along  with  any  of  the  Mendelian  genes  with  significant  correlation  with  the  cancer  gene  (rho  >  .2  for  p  <  .05/number  of  gene  pairs  tested).    Format:  Mendelian_gene_coexpressed_1  -­‐>  Cancer_gene_1(Cancer_gene_1  corrected  ranksum  p-­‐value).    Some  cancer  genes  have  no    Mendelian  gene  coexprssed  at  rho  >  .2,  but  the  set  of  Mendelian  genes  still  have  significantly  elevated  coexpression.  Format:    -­‐>  Cancer_gene_2(Cancer_gene_2  corrected  ranksum  p-­‐value).          

humannet_set  and  humannet:  "humannet_set"  shows  the  corrected  p-­‐value  for  the  connections  between  the  disease  pair's  genes  is  shown,  as  described  in  methods.  "humannet"  shows  all  connections.  Format:  Mendelian_gene_1  -­‐>  Cancer_gene_1,  Cancer_gene_2;  Mendelian_gene_2  -­‐>  Cancer_gene_3  …          

biogrid_set  and  biogrid:  identical  to  the  humannet  columns,  but  performed  on  the  BioGRID  network    

Page 132: Rachel D. Melamed

121

 

MD C gene_enrichmentgeneIntersection pathway_correlationpathway coex_CG coexpression humannet_sethumannet biogrid_set biogridChronic(Granulomatous(Disease BLCA 1 1 0.078547322 NCF4(@>(BCL2L1(1.69e@02);NCF2,NCF4,CYBB,CYBA(@>(DIAPH2(1.47e@02);NCF2,NCF4,CYBB,CYBA(@>(AHR(1.24e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.47e@02);NCF4,CYBA(@>(LRP5L(2.06e@02);NCF2,NCF4,CYBB,CYBA(@>(UBXN11(1.91e@02);NCF2,NCF4,CYBA(@>(ZNF586(4.38e@02);NCF2,NCF4,CYBB,CYBA(@>(SH3BGRL3(1.68e@02);NCF2,NCF4,CYBB,CYBA(@>(LPAR6(2.13e@02);CYBA(@>(TMEM80(2.68e@02);NCF2,NCF4,CYBB,CYBA(@>(KDM6A(1.51e@02);NCF2,NCF4,CYBB(@>(TGFBR1(1.43e@02);NCF2,NCF4,CYBB,CYBA(@>(IRF7(2.53e@02);NCF4,CYBB,CYBA(@>(CPM(1.31e@02);NCF4,CYBB,CYBA(@>(CD52(1.56e@02);NCF2,NCF4,CYBB(@>(RBM5(2.82e@02);NCF2,NCF4,CYBB,CYBA(@>(SPCS3(4.68e@02);NCF2,NCF4,CYBB,CYBA(@>(ARL8A(1.69e@02);NCF2,NCF4,CYBB,CYBA(@>(ARHGAP30(1.41e@02);NCF4(@>(PTPN7(3.47e@02);NCF2,NCF4,CYBB,CYBA(@>(CREBBP(1.41e@02);NCF2,NCF4(@>(KAT6A(3.73e@02)1 1Congenital(Ichthyosis BLCA 1 1 0.001525602 ALDH3A2,ABCA12(@>(FOXQ1(2.31e@02);CSTA,NIPAL4,LIPN(@>(AHR(2.31e@02);SPINK5,CSTA,KRT2,ABCA12,TGM1(@>(PVRL4(6.32e@04);ALOX12B,SPINK5,CSTA,ABCA12,TGM1(@>(LCE3D(5.41e@05);ALOX12B,SPINK5,CSTA,ABCA12,TGM1(@>(LCE3E(5.42e@05);ALOX12B,CSTA,ABCA12(@>(LCE3C(4.15e@04);ABCA12,TGM1(@>(EGFR(2.21e@02)1 1Polycystic(Kidney,(Autosomal(Dominant BLCA 1 0.096035489 TSC2(@>(Direct(p53(effectors(BCL2L1,TP53,EGFR,PTEN,AFP,CREBBP)0.77024073 1 1Diamond@Blackfan(Anemia BLCA 1 0.383933529 0.006200982 RPS26(@>(CDKN2A(8.89e@03);(@>(AMMECR1(4.13e@03);RPS19,RPL35A(@>(PABPC4(5.28e@04);RPS26,RPS19,RPS7(@>(TP53(1.79e@03);RPS26,RPS7(@>(UBE2T(4.93e@03);RPS26,RPS7(@>(PFDN2(1.00e@02);RPS26,RPS7(@>(TIMM17A(2.27e@03);RPS26(@>(POU5F1B(3.75e@04)1 1Inherited(Anomalies(of(the(Skin BLCA 1 1 0.013714977 DKC1(@>(CDKN2A(4.78e@02);NOP10(@>(PABPC4(2.54e@02);KRT6A,NHP2,KRT16(@>(EPS8L2(5.00e@02);WRAP53,DKC1,NHP2(@>(TP53(5.00e@02);KRT6C,KRT6A,KRT1,KRT16(@>(LCE3D(4.68e@02);KRT6C,KRT6A,KRT1,KRT9,KRT16(@>(LCE3E(2.54e@02);KRT6C,KRT6A,KRT9,KRT16(@>(LCE3C(1.00e@03);KRT6A,KRT9,KRT16(@>(EGFR(4.70e@02)1 1Spinocerebellar(Ataxia BLCA 1 0.151828695 0.000158829 ATXN7,ATM,TBP(@>(ORAOV1(3.31e@06);ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,PPP2R2B(@>(FHIT(6.28e@03);ZNF592,ATXN7,ATXN2,SYNE1,ATM,TTBK2,TBP,ITPR1,AFG3L2,SETX,PPP2R2B,ATXN1(@>(TNRC6A(1.69e@03);(@>(CHRFAM7A(2.37e@02);JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1(@>(CCSER1(2.37e@03);JPH3,CACNA1A,ATXN2,PDYN,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@>(DEAF1(2.43e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(FGFR3(5.62e@03);ZNF592,ATXN7,SYNE1,ATM,TTBK2,ITPR1,SETX(@>(PDE4D(4.83e@02);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(NOVA1(2.41e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(KLHDC9(3.90e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(OPCML(2.99e@03);ZNF592,POLG,ATXN2,TTBK2,TBP(@>(PHRF1(9.19e@03);JPH3,CACNA1A,ATXN2,PDYN,SPTBN2,TTBK2,KC����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Hypopituitarism BLCA 1 0.098019822 FGFR1(@>(Syndecan@3@mediated(signaling(events(EGFR,FGFR3);POU1F1(@>(Glucocorticoid(receptor(regulatory(network(TP53,AFP,CREBBP)0.552891743 0.53075862 1Combined(Heart(and(Skeletal(Defects BLCA 0.53 CREBBP 2.11E@27 CREBBP(@>(inhibition(of(huntingtons(disease(neurodegeneration(by(histone(deacetylase(inhibitors(CREBBP);CREBBP(@>(the(information(processing(pathway(at(the(ifn(beta(enhancer(IRF7,CREBBP);EP300,CREBBP(@>(Direct(p53(effectors(BCL2L1,TP53,EGFR,PTEN,AFP,CREBBP);EP300,CREBBP(@>(p53(pathway(CDKN2A,TP53,CREBBP);EP300(@>(Validated(transcriptional(targets(of(AP1(family(members(Fra1(and(Fra2(CCND1,CDKN2A);EP300,CREBBP(@>(acetylation(and(deacetylation(of(rela(in(nucleus(CREBBP);CREBBP(@>(wnt(signaling(pathway(CCND1,CREBBP);CREBBP(@>(Notch@HLH(transcription(pathway(CREBBP);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53,CREBBP);CREBBP(@>(Presenilin(action(in(Notch(and(Wnt(signaling(CCND1,CREBBP);EP300,CREBBP(@>(Glucocorticoid(receptor(regulatory(network(TP53,AFP,CREBBP);CREBBP(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,CREBBP);EP300,CREBBP(@>(Signaling(events(mediated(by(HDAC(Class(III(TP53,CREBBP);CREBBP(@>(Signaling(events(mediated(by(TCPTP(EGFR,CREBBP);EP300,CREBBP(@>(FOXM1(tra����������������������������������������������1 0.02531579 CREBBP(@>(TP53,TP53;EP300(@>(CREBBP0.53565909Specified(Hamartoses BLCA 0.62 PTEN 0.028031166 PTEN(@>(Direct(p53(effectors(BCL2L1,TP53,EGFR,PTEN,AFP,CREBBP);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);PTEN(@>(Negative(regulation(of(the(PI3K/AKT(network(PTEN);PTEN(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,CREBBP)0.739774885 0 PTEN(@>(EGFR;STK11(@>(TGFBR11Lipoprotein(Deficiencies BLCA 1 1 0.056770262 APOB,APOA1(@>(AFP(9.01e@03);(@>(UNC93A(1.46e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(AFM(6.27e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(CDHR5(6.41e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ALB(6.27e@03)1 1Disorders(of(Urea(Cycle(Metabolism BLCA 1 1 0.078763609 (@>(UNC93A(1.30e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(AFM(1.76e@02);NAGS,ARG1,ASL(@>(CDHR5(1.30e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ALB(1.76e@02)1 1Androgen(Insensitivity(Syndrome BRCA 1 0.002753113 AR(@>(Nongenotropic(Androgen(signaling(AKT1,PIK3R1,PIK3CA);AR(@>(FOXA1(transcription(factor(network(FOXA1,CDKN1B,NCOA3);AR(@>(Coregulation(of(Androgen(receptor(activity(CCND1,AKT1,CDKN2A,CASP8);AR(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDKN2A,TBL1XR1,CCND1,MED12,CDH1,TERT)1 1 0.2405 AR(@>(NCOA3Chronic(Granulomatous(Disease BRCA 1 0.407470815 0.076994822 NCF2,NCF4,CYBB,CYBA(@>(RPGR(1.37e@02);NCF2,NCF4,CYBB,CYBA(@>(SELPLG(1.16e@02);NCF2,NCF4,CYBA(@>(GPS2(1.52e@02);NCF2,NCF4,CYBB(@>(VPS9D1(1.65e@02);NCF2,NCF4,CYBB,CYBA(@>(MAP3K1(1.13e@02);NCF4(@>(CCDC18(3.95e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.24e@02);NCF2,NCF4,CYBB,CYBA(@>(ZNF276(1.19e@02);NCF2,CYBB,CYBA(@>(UBC(1.22e@02);NCF2,NCF4,CYBB,CYBA(@>(KDM5A(2.12e@02);NCF2,NCF4,CYBB,CYBA(@>(NFATC1(1.34e@02);NCF4,CYBA(@>(CDKN1B(4.01e@02);NCF2,NCF4,CYBB,CYBA(@>(HLA@A(1.79e@02);NCF2,NCF4,CYBB,CYBA(@>(NR1H2(1.11e@02);(@>(SPATA12(3.99e@02);NCF2,NCF4,CYBB,CYBA(@>(TICAM1(1.07e@02);NCF2,NCF4,CYBB,CYBA(@>(ITPR1(4.30e@02);NCF2,NCF4,CYBB(@>(MEF2A(3.00e@02);NCF2,NCF4,CYBB,CYBA(@>(RBM23(1.15e@02);NCF2,NCF4,CYBB,CYBA(@>(CTDP1(1.09e@02);CYBA(@>(NCOA3(1.67e@02);NCF2,NCF4,CYBB,CYBA(@>(RUNX1(1.11e@02);NCF2,NCF4,CYBB,CYBA(@>(TCF25(1.79e@02);NCF2,NCF4,CYBB,CYBA(@>(PQLC1(1.12e@02);NCF2,CYBB,CYBA(@>(NEU4(1.24e@02);NCF2,NCF4,CYBB,CYBA(@>(SRXN1(1.12e@02);NCF2,NCF4,CYBB,CYBA(@>(LPAR6(1.44e@02);NCF2,NCF4,CYBB,CYBA(@>(RBM7(1.10e@02);NCF2,NCF4,CYB���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Cerebral(Degeneration(Due(to(Generalized(LipidosesBRCA 1 1.93E@05 SMPD1(@>(Ceramide(signaling(pathway(PDGFA,MAP3K1,CASP8,AKT1,RB1,MAP2K4);SMPD1(@>(IL2(signaling(events(mediated(by(PI3K(MYB,TERT,PIK3CA,PIK3R1,AKT1);SMPD1(@>(phospholipids(as(signalling(intermediaries(PDGFA,PIK3CA,PIK3R1,AKT1);SMPD1(@>(ceramide(signaling(pathway(MAP3K1,CASP8,MAP2K4)0.159853943 1 1Congenital(Ichthyosis BRCA 1 1 0.073521831 ALOX12B,ALOXE3,CSTA,TGM1(@>(PARD6G(3.75e@02);ALOXE3,CSTA,ABCA12,TGM1(@>(CDH1(3.31e@02);SPINK5,CSTA,TGM1(@>(MUC21(2.98e@02);CSTA,NIPAL4,LIPN,ABHD5(@>(UBC(2.59e@02);ALOXE3,CSTA,NIPAL4(@>(PHLDA1(3.75e@02);ALOX12B(@>(KRTAP9@9(9.81e@03);ALOX12B(@>(KRTAP4@5(9.81e@03);CSTA,NIPAL4,LIPN,ABHD5(@>(TICAM1(3.83e@02);LIPN,ABHD5(@>(ZFP36L1(2.60e@02);CSTA,NIPAL4,LIPN,ABHD5(@>(SRXN1(2.98e@02)1 1Diamond@Blackfan(Anemia BRCA 1 0.968925499 0.001684618 (@>(MYB(3.65e@02);RPS26(@>(CDKN2A(1.09e@02);RPS26,RPS19,RPS10,RPL11,RPL35A,RPS7(@>(SLC25A5(1.93e@04);(@>(E2F4(4.16e@03);RPS26,RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7(@>(RPL13(6.67e@05);RPS26,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7(@>(RPL18(7.97e@05);(@>(NFE2L3(1.40e@03);RPS26,RPS19,RPS10,RPS7(@>(TCF3(2.93e@03);RPS26,RPS19,RPS7(@>(TP53(1.91e@03);RPL11,RPL35A,RPS7(@>(RBMX(2.97e@02);RPS26,RPS7(@>(FANCA(7.27e@03);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(TXNL4A(3.48e@02);(@>(CPNE7(2.04e@02);(@>(TERT(1.31e@02);RPS26,RPS19,RPS7(@>(HIST1H3B(7.83e@03)1 1Inherited(Anomalies(of(the(Skin BRCA 1 TERT 6.65E@05 ATP2A2(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,PIK3CA,PIK3R1,AKT1);TERT(@>(IL2(signaling(events(mediated(by(PI3K(MYB,TERT,PIK3CA,PIK3R1,AKT1);TERT(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(TERT,TP53);TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(AKT1,TERT,TP53,RB1);KRT1,TERT(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDKN2A,TBL1XR1,CCND1,MED12,CDH1,TERT);TERT(@>(role(of(nicotinic(acetylcholine(receptors(in(the(regulation(of(apoptosis(PIK3CA,TERT,PIK3R1,AKT1);TINF2,TERT,DKC1(@>(Regulation(of(Telomerase(CCND1,CDKN1B,TERT,AKT1)0.183228108 1 1Spinocerebellar(Ataxia BRCA 1 ITPR1 0.097855772 PRKCG(@>(EGFR(Inhibitor(Pathway,(Pharmacodynamics(ERBB2,PIK3CA,MAP3K1,PIK3R1,AKT1);ATM(@>(apoptotic(signaling(in(response(to(dna(damage(AKT1,TP53);ATM(@>(role(of(brca1(brca2(and(atr(in(cancer(susceptibility(TP53,FANCA);CACNA1A(@>(Anti@diabetic(Drug(Potassium(Channel(Inhibitors(Pathway,(Pharmacodynamics(AKT1,HNF1A,PIK3R1,PIK3CA);TBP(@>(Validated(targets(of(C@MYC(transcriptional(repression(ERBB2,CDKN1B,CCND1,ZFP36L1);ATM(@>(rb(tumor(suppressor/checkpoint(signaling(in(response(to(dna(damage(TP53,RB1);ATM(@>(E2F(transcription(factor(network(CDKN1B,CDKN2A,MCL1,RB1,E2F4);CACNA1A(@>(rac1(cell(motility(signaling(pathway(PIK3CA,MAP3K1,PIK3R1);ATM(@>(hypoxia(and(p53(in(the(cardiovascular(system(AKT1,TP53);TBP(@>(Glucocorticoid(receptor(regulatory(network(NFATC1,TP53,GATA3,AKT1);ATM(@>(Regulation(of(Telomerase(CCND1,CDKN1B,TERT,AKT1);PRKCG(@>(Retinoic(acid(receptors@mediated(signaling(AKT1,NCOR2,NCOA3);ATM(@>(Validated(transcriptional(targets(of(deltaNp63(isoforms(CDKN2A,RUNX1,AXL)0.031145236 APTX,SYT14,PPP2R2B(@>(LYRM2(2.24e@02);JPH3,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,PRKCG,FGF14,SYT14,PPP2R2B(@>(CXXC11(7.79e@03);(@>(CHRFAM7A(3.35e@02);ZNF592,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,SETX,ATXN1(@>(PIK3CA(2.81e@02);JPH3,ZNF592,ATXN7,ATXN2,TDP1,SYNE1,ATM,TTBK2,TBP,KCNC3,FGF14,AFG3L2,PPP2R2B(@>(MAP3K4(3.34e@03);JPH3,ATXN2,SYNE1,TTBK2,SYT14,PPP2R2B(@>(IGF1R(2.05e@02);JPH3,ZNF592,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,PRKCG,SETX,PPP2R2B,ATXN1(@>(WNK1(7.91e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,SYT14,PPP2R2B(@>(KCNN3(8.91e@03);ZNF592,ATXN7,ATXN2,TDP1,ATM,TTBK2,TBP,NOP56,AFG3L2,SETX,C10orf2(@>(CTCF(3.07e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@>(ASPHD1(1.05e@02);JPH3,CACNA1A,ATXN2,PDYN,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,SYT14,C10orf2,PPP2R2B(@>(GPR19(1.66e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,PRKCG,FGF14,SYT14,PPP2R2B(@>(IQSEC3(7.97e@03);JPH3,CACNA1A,ATXN2,PDYN,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Hypopituitarism BRCA 1 0.008828178 GH1(@>(regulation(of(eif@4e(and(p70s6(kinase(PTEN,AKT1,PIK3R1,PIK3CA);GLI2(@>(Signaling(events(mediated(by(the(Hedgehog(family(AKT1,PIK3R1,PIK3CA);GH1(@>(mtor(signaling(pathway(AKT1,PIK3CA,PIK3R1,PTEN);BTK(@>(Fc@epsilon(receptor(I(signaling(in(mast(cells(AKT1,MAP3K1,PIK3CA,PIK3R1,MAP2K4);BTK(@>(BCR(signaling(pathway(NFATC1,AKT1,PIK3CA,MAP3K1,PTEN,PIK3R1);GH1(@>(growth(hormone(signaling(pathway(PIK3CA,HNF1A,PIK3R1);POU1F1(@>(Glucocorticoid(receptor(regulatory(network(NFATC1,TP53,GATA3,AKT1);GH1(@>(akt(signaling(pathway(PIK3CA,PIK3R1,AKT1);GH1(@>(trefoil(factors(initiate(mucosal(healing(ERBB2,AKT1,MUC2,PIK3R1,PIK3CA);FGFR1(@>(FGF(signaling(pathway(CDH1,AKT1,FGFR2,PIK3R1,PIK3CA)1 1 1Combined(Heart(and(Skeletal(Defects BRCA 1 2.37E@10 CREBBP(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,PIK3CA,PIK3R1,AKT1);EP300,CREBBP(@>(IFN@gamma(pathway(AKT1,MAP3K1,PIK3R1,PIK3CA);EP300,CREBBP(@>(FOXA1(transcription(factor(network(FOXA1,CDKN1B,NCOA3);EP300,CREBBP(@>(transcription(regulation(by(methyltransferase(of(carm1(NCOA3,PRKAR1B);EP300(@>(Validated(transcriptional(targets(of(AP1(family(members(Fra1(and(Fra2(CCND1,NFATC1,CDKN2A);EP300(@>(Notch(signaling(pathway(NCOR2,CCND1,NOTCH3,MFAP2,GATA3,NCOR1);EP300(@>(Validated(nuclear(estrogen(receptor(alpha(network(CCND1,NCOA3,NCOR2,NCOR1);EP300(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDKN2A,TBL1XR1,CCND1,MED12,CDH1,TERT);EP300,CREBBP(@>(mechanism(of(gene(regulation(by(peroxisome(proliferators(via(ppara(FAT1,PRKAR1B,RB1,NCOR1,NCOR2);EP300(@>(Validated(targets(of(C@MYC(transcriptional(repression(ERBB2,CDKN1B,CCND1,ZFP36L1);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53,RB1);EP300,CREBBP(@>(E2F(transcription(factor(network(CDKN1B,CDKN2A,MCL1,RB1,E2F4);EP300,CRE������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.574890625 0.02290476 CREBBP(@>(NCOA3;EP300(@>(TP53;TBX5(@>(HNF1A,TP53,TBX30.2405 CREBBP(@>(NCOA3;EP300(@>(TP53Hereditary(Sensory(Neuropathy BRCA 1 0.035221138 NTRK1(@>(Trk(receptor(signaling(mediated(by(PI3K(and(PLC@gamma(CCND1,PIK3CA,PIK3R1,AKT1);HSPB1(@>(p38(mapk(signaling(pathway(MEF2A,MAP3K1,MAP2K4);HSPB1(@>(downregulated(of(mta@3(in(er@negative(breast(tumors(CDH1,MBD3);LMNA(@>(tnfr1(signaling(pathway(CASP8,MAP2K4);MED25(@>(Generic(Transcription(Pathway(MED12,MED15);NTRK1(@>(trka(receptor(signaling(pathway(PIK3CA,PIK3R1,AKT1);NDRG1(@>(Validated(targets(of(C@MYC(transcriptional(repression(ERBB2,CDKN1B,CCND1,ZFP36L1);NTRK1(@>(role(of(erk5(in(neuronal(survival(pathway(PIK3CA,MEF2A,PIK3R1,AKT1);EGR2(@>(IL4@mediated(signaling(events(AKT1,MYB,PIK3R1,PIK3CA);NTRK1(@>(p75(NTR)@mediated(signaling(AKT1,NDNL2,TP53,PIK3R1,PIK3CA);PMP22(@>(a6b1(and(a6b4(Integrin(signaling(CDH1,AKT1,ERBB2,PIK3R1,PIK3CA)0.403534273 1 1Severe(Combined(Immunodeficiency BRCA 1 0.033379032 CD3D,PTPRC,ZAP70(@>(t(cell(receptor(signaling(pathway(NFATC1,MAP3K1,PIK3CA,PIK3R1,MAP2K4);JAK3,IL2RG(@>(IL2(signaling(events(mediated(by(PI3K(MYB,TERT,PIK3CA,PIK3R1,AKT1);PTPRC(@>(BCR(signaling(pathway(NFATC1,AKT1,PIK3CA,MAP3K1,PTEN,PIK3R1);CD3D(@>(the(co@stimulatory(signal(during(t@cell(activation(PIK3CA,PIK3R1);JAK3(@>(CD40/CD40L(signaling(AKT1,MAP3K1,PIK3CA,PIK3R1,MAP2K4);JAK3,IL2RG(@>(il@7(signal(transduction(PIK3CA,PIK3R1);JAK3,IL2RG(@>(IL4@mediated(signaling(events(AKT1,MYB,PIK3R1,PIK3CA);CD3D(@>(Immunoregulatory(interactions(between(a(Lymphoid(and(a(non@Lymphoid(cell(CDH1,HLA@B,HLA@A);ADA(@>(Validated(transcriptional(targets(of(deltaNp63(isoforms(CDKN2A,RUNX1,AXL)0.001276652 ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(SELPLG(1.40e@03);ZAP70,IL2RG,JAK3,PTPRC,DCLRE1C(@>(ZNF384(5.80e@03);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(MAP3K1(3.39e@03);IL2RG,ADA,RFXAP,DCLRE1C(@>(CBFB(2.74e@04);RFXAP,AK2,DCLRE1C(@>(MYB(1.73e@04);IL2RG,ADA,RFXAP,PNP,RFX5,AK2,DCLRE1C(@>(CCDC18(1.15e@04);RFXAP,DCLRE1C(@>(FNTA(4.47e@02);(@>(E2F4(9.86e@04);ZAP70,RFXANK,IL7R,CD3D(@>(RPL13(3.43e@03);NHEJ1,RFXANK(@>(RPL18(5.22e@03);(@>(GATA3(1.97e@02);ZAP70,IL2RG,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D(@>(USP36(1.43e@02);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(ZNF276(3.74e@03);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(KDM5A(9.26e@03);ZAP70,IL2RG,JAK3,RFXAP,CD3D,DCLRE1C(@>(CTCF(2.54e@02);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC(@>(NFATC1(1.90e@02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(CDKN1B(1.09e@02);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(HLA@A(7.19e@03);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(DENND4B(3.72e@03);ZAP70,IL2RG,RFXAP,IL��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Specified(Hamartoses BRCA 0.7 PTEN 0.000162829 PTEN(@>(skeletal(muscle(hypertrophy(is(regulated(via(akt@mtor(pathway(PTEN,PIK3CA,IGF1R,PIK3R1,AKT1);PTEN(@>(regulation(of(eif@4e(and(p70s6(kinase(PTEN,AKT1,PIK3R1,PIK3CA);VHL(@>(vegf(hypoxia(and(angiogenesis(PIK3CA,PIK3R1,AKT1);PTEN(@>(mtor(signaling(pathway(AKT1,PIK3CA,PIK3R1,PTEN);PTEN(@>(BCR(signaling(pathway(NFATC1,AKT1,PIK3CA,MAP3K1,PTEN,PIK3R1);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);PTEN(@>(RhoA(signaling(pathway(PTEN,CDKN1B,MAP2K4);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN,AKT1,PIK3CA,PIK3R1,CDKN1B);PTEN(@>(Negative(regulation(of(the(PI3K/AKT(network(PTEN,AKT1);PTEN(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,PIK3CA,PIK3R1,AKT1)0.736815818 0 PTEN(@>(IGF1R;SDHB(@>(CDKN2A;STK11(@>(TP531Li(Fraumeni(and(Related(Syndromes BRCA 0.05 CDKN2A,TP53 8.57E@15 TP53(@>(chaperones(modulate(interferon(signaling(pathway(TP53,RB1);TP53(@>(apoptotic(signaling(in(response(to(dna(damage(AKT1,TP53);CDKN2A(@>(Validated(transcriptional(targets(of(AP1(family(members(Fra1(and(Fra2(CCND1,NFATC1,CDKN2A);TP53(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(TERT,TP53);CDKN2A(@>(Coregulation(of(Androgen(receptor(activity(CCND1,AKT1,CDKN2A,CASP8);TP53,CHEK2(@>(role(of(brca1(brca2(and(atr(in(cancer(susceptibility(TP53,FANCA);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(AKT1,TERT,TP53,RB1);CDKN2A(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDKN2A,TBL1XR1,CCND1,MED12,CDH1,TERT);TP53(@>(btg(family(proteins(and(cell(cycle(regulation(CCND1,TP53,RB1);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53(@>(p53(signaling(pathway(CCND1,TP53,RB1);TP53(@>(regulation(of(transcriptional(activity(by(pml(TP53,RB��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.720244363 1 1Lipoprotein(Deficiencies BRCA 1 0.890189985 0.077804012 (@>(KCNMB3(1.34e@02);(@>(UNC93A(1.98e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(F10(1.16e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(CYP2E1(1.16e@02);APOB,LCAT,APOA1(@>(SLC6A12(1.84e@02);MTTP,APOB,LCAT,SAR1B,APOA1,ABCA1(@>(NEU4(1.45e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(HNF1A(1.45e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(AQP11(2.84e@02)1 1Disorders(of(Urea(Cycle(Metabolism BRCA 1 0.223337184 0.086959582 (@>(TMEM184A(4.36e@02);(@>(KCNMB3(1.76e@02);(@>(FOXA1(2.92e@02);ASS1,NAGS,ASL,CPS1(@>(HSBP1L1(1.85e@02);(@>(UNC93A(1.76e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(F10(3.36e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(CYP2E1(3.36e@02);ASS1(@>(TBX3(4.36e@02);NAGS,ARG1,ASL,CPS1(@>(SLC6A12(2.79e@02);ASS1,NAGS,ARG1,ASL(@>(ISOC2(2.57e@02);NAGS,ASL(@>(CDK10(4.25e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(HNF1A(1.49e@02);NAGS,ARG1(@>(AQP11(2.08e@02)1 1Retinitis(Pigmentosa BRCA 1 RPGR 1 1.07E@13 TTC8,CERKL,FAM161A(@>(CHRFAM7A(2.17e@02);SNRNP200,CA4,MERTK,PDE6B,PRPF3,BEST1,RP2,TOPORS,FAM161A,SEMA4A(@>(VEZF1(1.76e@03);CA4,KLHL7,SPATA7,CRB1,FAM161A(@>(IGF1R(2.39e@02);CNGA1(@>(MUC20(4.95e@03);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(SLC6A13(4.54e@13);CRX,LRAT,FSCN2,RDH12,SNRNP200,PRPH2,RHO,CNGB1,EYS,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(CROCC(8.93e@16);CRX,SNRNP200,TULP1(@>(CCDC144NL(8.18e@04)0.04182609 IMPDH1(@>(PABPC3;PRPF3(@>(RPL18;PRPF31(@>(RPL13,TXNL4A;PRPF8(@>(SF3B1,PABPC3;RHO(@>(SF3B1;SNRNP200(@>(TXNL4A,PABPC31Haemophilia BRCA 1 0.052116948 F9(@>(Formation(of(Fibrin(Clot((Clotting(Cascade)(F10)1 0.702 1Chronic(Granulomatous(Disease COAD 1 1 0.087493235 NCF2(@>(KRAS(4.54e@02);NCF2,CYBB,CYBA(@>(TCF7L2(1.51e@02);NCF2,NCF4,CYBB,CYBA(@>(PCBP1(1.53e@02);NCF2,NCF4,CYBB,CYBA(@>(FBRS(1.53e@02);NCF2,NCF4(@>(TNFRSF10C(1.54e@02);NCF2,NCF4,CYBB(@>(TRAPPC11(4.57e@02);NCF2,CYBB(@>(ACVR2A(4.65e@02);NCF2,NCF4,CYBB,CYBA(@>(GGT1(4.40e@02);NCF2,NCF4,CYBB,CYBA(@>(BRAF(2.90e@02)1 1Polycystic(Kidney,(Autosomal(Dominant COAD 1 1.58E@05 TSC2(@>(LKB1(signaling(events(SMAD4,TP53);TSC2(@>(mTOR(signaling(pathway(NRAS,BRAF,KRAS);TSC2(@>(Validated(targets(of(C@MYC(transcriptional(repression(SMAD4,SMAD2,SMAD3)0.807669883 1 0.26722222Inherited(Anomalies(of(the(Skin COAD 1 2.64E@06 TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(TP53,KRAS);KRT1,TERT(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(APC,CDKN2A,TCF7L2);TERT(@>(Validated(targets(of(C@MYC(transcriptional(activation(SMAD4,TP53,SMAD3)0.147886703 1 1Spinocerebellar(Ataxia COAD 1 0.380460894 0.003170856 ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,PPP2R2B(@>(FHIT(3.79e@03);JPH3,CACNA1A,ATXN2,SYNE1,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(PLK5(2.24e@04);ZNF592,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,ATXN1(@>(PIK3CA(8.50e@03);JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1(@>(CCSER1(1.43e@03);JPH3,ZNF592,CACNA1A,ATXN7,ATXN2,PDYN,TDP1,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B,ATXN1(@>(WHSC1L1(2.24e@04);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2B(@>(PARK2(1.71e@03);JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B,ATXN1(@>(ZC3H13(6.35e@04);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(DLGAP1(1.67e@03);ATM(@>(FNTA(3.08e@02);JPH3,ATXN2,SPTBN2,TTBK2,KCNC3,PPP2R2B(@>(ENPP6(7.78e@03);JPH3,APTX,CACNA1A,ATXN7,ATXN2,PDYN,TDP1,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,PRNP,ITPR1,PRKCG,�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Combined(Heart(and(Skeletal(Defects COAD 1 0.00325544 EP300(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(APC,CDKN2A,TCF7L2);EP300(@>(Validated(targets(of(C@MYC(transcriptional(repression(SMAD4,SMAD2,SMAD3);EP300,CREBBP(@>(Validated(targets(of(C@MYC(transcriptional(activation(SMAD4,TP53,SMAD3);CREBBP(@>(Signaling(events(mediated(by(TCPTP(PIK3CA,INS)0.664784536 0.05772 CREBBP(@>(TP53,UBE3A;EP300(@>(TP531Neurofibromatosis COAD 1 0.012213067 NF1(@>(Regulation(of(Ras(family(activation(NRAS,KRAS)1 0.55985246 0.53565909Hereditary(Sensory(Neuropathy COAD 1 0.038741758 NTRK1(@>(Trk(receptor(signaling(mediated(by(PI3K(and(PLC@gamma(PIK3CA,NRAS,KRAS);NTRK1(@>(ARMS@mediated(activation(BRAF);NDRG1(@>(Validated(targets(of(C@MYC(transcriptional(repression(SMAD4,SMAD2,SMAD3);NTRK1(@>(Frs2@mediated(activation(BRAF);NTRK1(@>(Signalling(to(p38(via(RIT(and(RIN(BRAF)0.718504577 1 1Severe(Combined(Immunodeficiency COAD 1 0.461439204 0.001993726 ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(B2M(8.02e@03);ZAP70,IL2RG,JAK3,RFXAP,IL7R,CD3D,DCLRE1C(@>(CSTF2T(1.17e@02);IL2RG,JAK3,RFXANK,RFX5,PTPRC(@>(PRR14(6.59e@03);IL2RG,JAK3,PNP,PTPRC(@>(PCBP1(4.79e@02);ZAP70,IL2RG,JAK3,PTPRC,CD3D,DCLRE1C(@>(SMAD2(7.76e@03);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(IKZF3(8.72e@05);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(ZNF18(8.78e@03);(@>(NRAS(4.87e@02)1 1Specified(Hamartoses COAD 1 0.129885453 0.682837057 0 PTEN(@>(PIK3CA;SDHB(@>(CDKN2A;STK11(@>(TP531Li(Fraumeni(and(Related(Syndromes COAD 0.03 CDKN2A,TP53 1.64E@21 CDKN2A(@>(C@MYC(pathway(CDKN2A,FBXW7);TP53(@>(LKB1(signaling(events(SMAD4,TP53);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(TP53,KRAS);CDKN2A(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(APC,CDKN2A,TCF7L2);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53,CHEK2(@>(PLK3(signaling(events(TP53);TP53(@>(Validated(targets(of(C@MYC(transcriptional(activation(SMAD4,TP53,SMAD3);CDKN2A(@>(Validated(transcriptional(targets(of(deltaNp63(isoforms(CDKN2A,FBXW7)0.700494843 0.1924 CHEK2(@>(TP53,SMAD4;TP53(@>(UBE3A1Retinitis(Pigmentosa COAD 1 0.968925499 0.001527057 CA4,SPATA7,CRB1,CERKL(@>(PLK5(1.76e@02);(@>(IMMP2L(5.09e@05);SPATA7,CRB1,MERTK,PDE6B,PRCD,FAM161A(@>(ENPP6(2.88e@03)1 0.64133333Hereditary(Hemorrhagic(Telangiectasia COAD 0.46 SMAD4 3.56E@24 SMAD4(@>(TGF@beta(receptor(signaling(SMAD4,SMAD2,SMAD3);SMAD4,ACVRL1,ENG(@>(ALK1(signaling(events(SMAD4,ACVR2A,ID1);SMAD4(@>(LKB1(signaling(events(SMAD4,TP53);SMAD4(@>(Signaling(by(BMP(SMAD4);SMAD4(@>(Signaling(by(TGF(beta(SMAD4);SMAD4(@>(Regulation(of(cytoplasmic(and(nuclear(SMAD2/3(signaling(SMAD4,SMAD2,SMAD3);SMAD4(@>(Validated(targets(of(C@MYC(transcriptional(repression(SMAD4,SMAD2,SMAD3);SMAD4(@>(Validated(targets(of(C@MYC(transcriptional(activation(SMAD4,TP53,SMAD3)0.518227346 0 SMAD4(@>(ACVR2A 1Dopa@Responsive(Dystonia COAD 0.53 TH 0.000139524 TH(@>(Alpha@synuclein(signaling(TH,PARK2) 1 1 1Chronic(Granulomatous(Disease GBM 1 1 0.07356062 NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.48e@02);NCF2,NCF4,CYBB,CYBA(@>(ZZEF1(1.72e@02);NCF4,CYBA(@>(CDKN1B(4.22e@02);(@>(C12orf5(1.72e@02);NCF2,NCF4,CYBB,CYBA(@>(HRH2(9.80e@03);NCF2(@>(TNFRSF9(1.98e@02);NCF2,NCF4,CYBB(@>(STAG2(2.52e@02);NCF2,NCF4,CYBB,CYBA(@>(PRDM2(4.38e@02);NCF2,NCF4,CYBB,CYBA(@>(RAP1B(1.23e@02);NCF2,CYBB,CYBA(@>(METTL9(1.10e@02);NCF2,NCF4,CYBB,CYBA(@>(LPAR6(1.32e@02);NCF2,NCF4,CYBB,CYBA(@>(CAMK1D(1.56e@02);NCF2,NCF4,CYBB,CYBA(@>(IGSF6(3.92e@02);NCF2,NCF4,CYBB,CYBA(@>(RB1(1.78e@02);NCF2,NCF4,CYBB,CYBA(@>(SCIMP(2.19e@02);NCF2,NCF4,CYBB,CYBA(@>(EVI2A(1.49e@02);NCF2,NCF4,CYBB,CYBA(@>(EVI2B(1.30e@02);NCF2,NCF4,CYBB,CYBA(@>(BRAF(2.52e@02)1 1Cerebral(Degeneration(Due(to(Generalized(LipidosesGBM 1 0.039068638 SMPD1(@>(IL2(signaling(events(mediated(by(PI3K(AKT1,PIK3R1,PIK3CA);SMPD1(@>(phospholipids(as(signalling(intermediaries(AKT1,PDGFRA,PIK3R1,PIK3CA)0.165898202 1 1Congenital(Ichthyosis GBM 1 0.741625429 0.024789969 (@>(CDKN2B(2.27e@03);ALOX12B,CSTA(@>(ADSSL1(4.30e@02);SPINK5,CSTA(@>(SVIL(3.26e@02);ABCA12,TGM1(@>(EGFR(4.30e@02)0.8658 1Inherited(Adrenogenital(Disorders GBM 1 0.521588572 0.097166245 HSD3B2,POR,CYP17A1,CYP21A2(@>(ITIH1(1.83e@02)1 1Pervasive,(Specified(Congenital(Anomalies GBM 1 BRAF 2.63E@31 NRAS,HRAS,RAF1,SOS1,KRAS(@>(EGFR(Inhibitor(Pathway,(Pharmacodynamics(PIK3CA,EGFR,CAMK1D,PIK3R1,AKT1);HRAS,SOS1(@>(nfat(and(hypertrophy(of(the(heart((AKT1,PIK3R1,PIK3CA);HRAS,NRAS,SOS1,KRAS(@>(Trk(receptor(signaling(mediated(by(PI3K(and(PLC@gamma(PIK3CA,PIK3R1,AKT1);HRAS,RAF1,SOS1(@>(pdgf(signaling(pathway(PIK3CA,PDGFRA,PIK3R1);PTCH1(@>(Signaling(events(mediated(by(the(Hedgehog(family(AKT1,PIK3R1,PIK3CA);SOS1(@>(PDGFR@alpha(signaling(pathway(PIK3CA,PDGFRA,PIK3R1);HRAS,SOS1(@>(vegf(hypoxia(and(angiogenesis(PIK3CA,PIK3R1,AKT1);PTPN11(@>(IFN@gamma(pathway(AKT1,RAP1B,PIK3R1,PIK3CA);PTPN11,SOS1(@>(IL2(signaling(events(mediated(by(STAT5(PIK3CA,CCND2,CDK6,PIK3R1);PTPN11(@>(VEGFR1(specific(signals(AKT1,PIK3R1,PIK3CA);HRAS(@>(LPA(receptor(mediated(events(AKT1,EGFR,PIK3R1);HRAS,RAF1(@>(Nongenotropic(Androgen(signaling(AKT1,PIK3R1,PIK3CA);PTPN11,SOS1(@>(IL2(signaling(events(mediated(by(PI3K(AKT1,PIK3R1,PIK3CA);HRAS,RAF1,PTPN11,SOS1(@>(Fc@epsilon(receptor(I(signaling(in(mast(cells(AKT1,PIK3R1,PIK3CA);HRAS,RAF1,SOS1(@>(tpo�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.115677047 0 BRAF(@>(EGFR,DGKZ;HRAS(@>(BRAF,EGFR;KRAS(@>(RAP1B;NRAS(@>(BRAF,PIK3CA;PTPN11(@>(PTEN,EGFR,RB1,STAG2,STAG20.2405 BRAF(@>(AKT1,ARID2,TP53,DGKZ;CHD7(@>(TP53;CUL7(@>(NF1;HRAS(@>(PIK3CA;KRAS(@>(BRAF;NIPBL(@>(PIK3R1,PIK3CA,CDK6,MDM2,CD33;NRAS(@>(EGFR;PTPN11(@>(PIK3R1,RB1;RAF1(@>(AKT1;RPS6KA3(@>(BRAF,EGFR,STAG2;SMC1A(@>(CDK4,STAG2;SMC3(@>(CDK4,MET;SOS1(@>(EGFR;TRIM32(@>(PIK3R1,MYCNDiamond@Blackfan(Anemia GBM 0.7 RPL5 1 0.011511663 RPS26,RPS19,RPS10,RPS7(@>(SIVA1(1.42e@03);RPS26,RPS24,RPS7(@>(TFB2M(2.01e@03);RPS26(@>(CDKN2A(1.64e@02);RPS24,RPS10,RPL11,RPL35A(@>(RPL5(7.91e@04);(@>(C12orf5(4.18e@02);(@>(CCNE1(1.34e@02);RPS26,RPS19,RPS7(@>(TP53(2.74e@03);RPS26(@>(CDK4(1.65e@02);RPS26,RPS19,RPL35A,RPS7(@>(CNIH4(2.94e@03)1 0.26054167Inherited(Anomalies(of(the(Skin GBM 1 1.50E@05 ATP2A2(@>(nfat(and(hypertrophy(of(the(heart((AKT1,PIK3R1,PIK3CA);TERT(@>(IL2(signaling(events(mediated(by(PI3K(AKT1,PIK3R1,PIK3CA);TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(AKT1,TP53,RB1);TERT(@>(role(of(nicotinic(acetylcholine(receptors(in(the(regulation(of(apoptosis(PIK3CA,PIK3R1,AKT1);TINF2,TERT,DKC1(@>(Regulation(of(Telomerase(CDKN1B,EGFR,AKT1)0.165069058 1 1Spinocerebellar(Ataxia GBM 1 0.007540264 PRKCG(@>(EGFR(Inhibitor(Pathway,(Pharmacodynamics(PIK3CA,EGFR,CAMK1D,PIK3R1,AKT1);ATM(@>(apoptotic(signaling(in(response(to(dna(damage(AKT1,TP53);ATM(@>(p53(pathway(AKT1,CDKN2A,RPL5,TP53,MDM2);ATM(@>(cell(cycle:(g2/m(checkpoint(MDM2,TP53);CACNA1A(@>(Anti@diabetic(Drug(Potassium(Channel(Inhibitors(Pathway,(Pharmacodynamics(AKT1,PIK3R1,PIK3CA);ATM(@>(BARD1(signaling(events(CCNE1,TP53);ATM(@>(atm(signaling(pathway(MDM2,TP53);ATM(@>(rb(tumor(suppressor/checkpoint(signaling(in(response(to(dna(damage(CDK4,TP53,RB1);ATM(@>(E2F(transcription(factor(network(CDKN1B,CDKN2A,CCNE1,RB1,CDKN2C);CACNA1A(@>(rac1(cell(motility(signaling(pathway(PIK3CA,PDGFRA,PIK3R1);ATM(@>(hypoxia(and(p53(in(the(cardiovascular(system(AKT1,MDM2,TP53);ATM(@>(Regulation(of(Telomerase(CDKN1B,EGFR,AKT1)0.003470279 POLG,ATXN7,TDP1,ATM,TBP,NOP56,AFG3L2,C10orf2(@>(POLR3E(2.52e@02);POLG,ATXN7,TDP1,ATM,TTBK2,TBP,ITPR1(@>(MARCH9(5.79e@03);ZNF592,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,ATXN1(@>(PIK3CA(9.40e@03);JPH3,APTX,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(NF1(1.85e@03);JPH3,ZNF592,ATXN2,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,PPP2R2B(@>(PIK3C2B(6.44e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(FGFR3(3.99e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(BRSK2(1.51e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(LSAMP(1.98e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(NETO1(1.98e@03);JPH3,CACNA1A,TDP1,SYNE1,ATM,TTBK2,TBP,KCNC3,ITPR1,FGF14(@>(ELP4(1.93e@03);TTBK2,AFG3L2(@>(UQCRC2(4.16e@02);JPH3,CACNA1A,ATXN7,TDP1,SYNE1,ATM,SPTBN2,TTBK2,T����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Hypopituitarism GBM 1 8.68E@11 GH1(@>(regulation(of(eif@4e(and(p70s6(kinase(PTEN,AKT1,PIK3R1,PIK3CA);GLI2(@>(Signaling(events(mediated(by(the(Hedgehog(family(AKT1,PIK3R1,PIK3CA);GH1(@>(mtor(signaling(pathway(AKT1,PIK3CA,PIK3R1,PTEN);BTK(@>(Fc@epsilon(receptor(I(signaling(in(mast(cells(AKT1,PIK3R1,PIK3CA);BTK(@>(BCR(signaling(pathway(AKT1,VAV2,PIK3CA,PIK3R1,PTEN);FGFR1(@>(Syndecan@3@mediated(signaling(events(EGFR,FGFR3);FGFR1(@>(Signal(transduction(by(L1(VAV2,EGFR);GH1(@>(growth(hormone(signaling(pathway(PIK3CA,PIK3R1);GH1(@>(akt(signaling(pathway(PIK3CA,PIK3R1,AKT1);BTK(@>(phosphoinositides(and(their(downstream(targets(AKT1,VAV2);GH1(@>(trefoil(factors(initiate(mucosal(healing(AKT1,EGFR,PIK3R1,PIK3CA);FGFR1(@>(FGF(signaling(pathway(AKT1,FGFR3,PIK3CA,PIK3R1,MET)1 0.26236364 1Combined(Heart(and(Skeletal(Defects GBM 1 6.38E@05 CREBBP(@>(nfat(and(hypertrophy(of(the(heart((AKT1,PIK3R1,PIK3CA);EP300,CREBBP(@>(IFN@gamma(pathway(AKT1,RAP1B,PIK3R1,PIK3CA);EP300,CREBBP(@>(Direct(p53(effectors(TP53,MET,EGFR,C12orf5,RB1,PTEN,MDM2);EP300,CREBBP(@>(p53(pathway(AKT1,CDKN2A,RPL5,TP53,MDM2);EP300(@>(cell(cycle:(g2/m(checkpoint(MDM2,TP53);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53,RB1);EP300,CREBBP(@>(E2F(transcription(factor(network(CDKN1B,CDKN2A,CCNE1,RB1,CDKN2C);EP300,CREBBP(@>(il@7(signal(transduction(PIK3CA,PIK3R1);EP300(@>(hypoxia(and(p53(in(the(cardiovascular(system(AKT1,MDM2,TP53);CREBBP(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,PIK3CA,AKT1,PIK3R1,PIK3C2B);EP300(@>(ATF@2(transcription(factor(network(PDGFRA,CDK4,NF1,RB1);CREBBP(@>(Signaling(events(mediated(by(TCPTP(MET,EGFR,PIK3R1,PIK3CA);EP300,CREBBP(@>(FOXM1(transcription(factor(network(CDKN2A,CDK4,CCNE1,RB1)0.662181163 0.10307143 CREBBP(@>(CTBP1;EP300(@>(TP53,TP531Specified(Anomalies(of(the(Musculoskeletal(SystemGBM 0.72 FGFR3 0.013251879 SNAI2(@>(Direct(p53(effectors(TP53,MET,EGFR,C12orf5,RB1,PTEN,MDM2);FGFR3(@>(Syndecan@3@mediated(signaling(events(EGFR,FGFR3);FGFR3(@>(FGFR3b(ligand(binding(and(activation(FGFR3);FGFR3(@>(FGFR3c(ligand(binding(and(activation(FGFR3);MITF,SNAI2(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,PIK3CA,AKT1,PIK3R1,PIK3C2B);MITF(@>(IL6@mediated(signaling(events(AKT1,PIK3R1,PIK3CA);FGFR3(@>(FGF(signaling(pathway(AKT1,FGFR3,PIK3CA,PIK3R1,MET)0.867886984 1 1Neurofibromatosis GBM 0.44 NF1 0.003812225 NF1(@>(ATF@2(transcription(factor(network(PDGFRA,CDK4,NF1,RB1)0.689368988 0 NF1(@>(EVI2A,CDKN2A;NF2(@>(NF11Hereditary(Sensory(Neuropathy GBM 1 INF2 0.097855772 NTRK1(@>(Trk(receptor(signaling(mediated(by(PI3K(and(PLC@gamma(PIK3CA,PIK3R1,AKT1);NDRG1(@>(Direct(p53(effectors(TP53,MET,EGFR,C12orf5,RB1,PTEN,MDM2);NTRK1(@>(trka(receptor(signaling(pathway(PIK3CA,PIK3R1,AKT1);NTRK1(@>(role(of(erk5(in(neuronal(survival(pathway(PIK3CA,PIK3R1,AKT1);NTRK1(@>(Signalling(to(p38(via(RIT(and(RIN(BRAF);EGR2(@>(IL4@mediated(signaling(events(AKT1,PIK3R1,PIK3CA);NTRK1(@>(p75(NTR)@mediated(signaling(AKT1,OMG,TP53,PIK3R1,PIK3CA);PMP22(@>(a6b1(and(a6b4(Integrin(signaling(AKT1,MET,EGFR,PIK3R1,PIK3CA)0.381135563 1 1Tuberous(Sclerosis GBM 1 0.004334094 TSC2(@>(AKT(phosphorylates(targets(in(the(cytosol(AKT1,MDM2);TSC2,TSC1(@>(mtor(signaling(pathway(AKT1,PIK3CA,PIK3R1,PTEN);TSC2(@>(Direct(p53(effectors(TP53,MET,EGFR,C12orf5,RB1,PTEN,MDM2);TSC2,TSC1(@>(mTOR(signaling(pathway(EEF2K,AKT1,CCNE1,BRAF)0.686334259 0.63004225 0.180375 TSC1(@>(CDKN1B;TSC2(@>(AKT1,CDKN1BSevere(Combined(Immunodeficiency GBM 1 0.000972189 JAK3,IL2RG(@>(IL2(signaling(events(mediated(by(STAT5(PIK3CA,CCND2,CDK6,PIK3R1);JAK3,IL2RG(@>(IL2(signaling(events(mediated(by(PI3K(AKT1,PIK3R1,PIK3CA);PTPRC(@>(BCR(signaling(pathway(AKT1,VAV2,PIK3CA,PIK3R1,PTEN);CD3D(@>(the(co@stimulatory(signal(during(t@cell(activation(PIK3CA,PIK3R1);JAK3(@>(CD40/CD40L(signaling(AKT1,PIK3R1,PIK3CA);JAK3,IL2RG(@>(il@7(signal(transduction(PIK3CA,PIK3R1);JAK3,IL2RG(@>(IL4@mediated(signaling(events(AKT1,PIK3R1,PIK3CA);JAK3(@>(Signaling(events(mediated(by(TCPTP(MET,EGFR,PIK3R1,PIK3CA);CD3D,PTPRC(@>(CXCR4@mediated(signaling(events(PTEN,AKT1,RAP1B,PIK3R1,PIK3CA)0.001527057 IL2RG,PNP,PTPRC,DCLRE1C(@>(CD33(4.61e@03);ZAP70,ADA,RFXANK,CD3D(@>(SIVA1(6.96e@03);ZAP70,IL2RG,JAK3,RFXAP,IL7R,CD3D(@>(POLR3E(1.28e@02);ZAP70,IL2RG,JAK3,RFXAP,RFX5,IL7R,CD3D,DCLRE1C(@>(MARCH9(1.12e@02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(ZZEF1(3.17e@02);ZAP70,RFXAP,IL7R,CD3D,DCLRE1C(@>(ELP4(3.17e@02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(CDKN1B(1.14e@02);ZAP70,IL2RG,NHEJ1,RFX5,IL7R,CD3D,DCLRE1C(@>(RPL5(5.40e@05);IL2RG,JAK3,PTPRC,DCLRE1C(@>(HRH2(3.56e@02);IL2RG,CIITA,PNP,RFX5(@>(TNFRSF9(1.40e@02);IL2RG,JAK3,ADA,RFXAP,PNP,PTPRC,DCLRE1C(@>(STAG2(1.27e@02);ZAP70,IL2RG,JAK3,RFXAP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1(@>(ARID2(1.14e@02);IL2RG,CIITA,ADA,PNP,PTPRC(@>(RAP1B(4.05e@02);ADA,AK2,DCLRE1C(@>(CDK6(3.58e@04);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C(@>(LPAR6(6.96e@03);IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,DCLRE1C(@>(RB1(3.56e@03);IL2RG,CIITA,RFX5,PTPRC,DCLRE1C(@>(SCIMP(4.22e@03);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(EVI2A(1.27e@02);IL2RG,JAK3,RFX5,PTPRC,DCLRE1C�������������������1 1Specified(Hamartoses GBM 0.62 PTEN 3.91E@09 PTEN(@>(skeletal(muscle(hypertrophy(is(regulated(via(akt@mtor(pathway(PTEN,PIK3CA,PIK3R1,AKT1);PTEN(@>(regulation(of(eif@4e(and(p70s6(kinase(PTEN,AKT1,PIK3R1,PIK3CA);VHL(@>(vegf(hypoxia(and(angiogenesis(PIK3CA,PIK3R1,AKT1);PTEN(@>(mtor(signaling(pathway(AKT1,PIK3CA,PIK3R1,PTEN);PTEN(@>(Direct(p53(effectors(TP53,MET,EGFR,C12orf5,RB1,PTEN,MDM2);PTEN(@>(BCR(signaling(pathway(AKT1,VAV2,PIK3CA,PIK3R1,PTEN);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN,AKT1,PIK3CA,PIK3R1,CDKN1B);PTEN(@>(Negative(regulation(of(the(PI3K/AKT(network(PTEN,AKT1);PTEN(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,PIK3CA,AKT1,PIK3R1,PIK3C2B);PTEN(@>(CXCR4@mediated(signaling(events(PTEN,AKT1,RAP1B,PIK3R1,PIK3CA)0.451482314 0 PTEN(@>(AKT1;SDHD(@>(EGFR;STK11(@>(CDKN2A1Li(Fraumeni(and(Related(Syndromes GBM 0.03 CDKN2A,TP53 4.83E@30 TP53(@>(Aurora(A(signaling(AKT1,MDM2,TP53);TP53(@>(chaperones(modulate(interferon(signaling(pathway(TP53,RB1);TP53(@>(apoptotic(signaling(in(response(to(dna(damage(AKT1,TP53);TP53(@>(Direct(p53(effectors(TP53,MET,EGFR,C12orf5,RB1,PTEN,MDM2);TP53,CDKN2A,CHEK2(@>(p53(pathway(AKT1,CDKN2A,RPL5,TP53,MDM2);TP53,CHEK2(@>(cell(cycle:(g2/m(checkpoint(MDM2,TP53);TP53(@>(estrogen(responsive(protein(efp(controls(cell(cycle(and(breast(tumors(growth(CDK4,TP53,CDK6);CDKN2A(@>(Coregulation(of(Androgen(receptor(activity(AKT1,CDKN2A,SVIL,CDK6);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(AKT1,TP53,RB1);TP53(@>(BARD1(signaling(events(CCNE1,TP53);TP53(@>(btg(family(proteins(and(cell(cycle(regulation(TP53,RB1);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53,CHEK2(@>(atm(signaling(pathway(MDM2,TP53);TP53,CHEK2(@>(PLK3(signaling(events(CCNE1,TP53);TP53(@>(p53(signaling(pathway(CDK4,MDM2,CCNE1,TP53,RB1);TP5������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.665220849 0 CDKN2A(@>(CDKN2C;CHEK2(@>(CCND2;TP53(@>(PTEN0.2405 CDKN2A(@>(MYCN;CHEK2(@>(TP53;TP53(@>(CCND2Lipoprotein(Deficiencies GBM 1 0.930390184 0.052743895 APOB,SAR1B(@>(IDH1(1.66e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ITIH1(5.60e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(DMRTA1(1.54e@02)1 1Disorders(of(Urea(Cycle(Metabolism GBM 1 0.086191281 ARG1(@>(IL4@mediated(signaling(events(AKT1,PIK3R1,PIK3CA);ARG1(@>(ATF@2(transcription(factor(network(PDGFRA,CDK4,NF1,RB1)0.093008554 NAGS,ASL(@>(VAV2(3.02e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ITIH1(1.71e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(DMRTA1(1.71e@02)1 1Retinitis(Pigmentosa GBM 1 0.439405616 3.62E@10 SNRNP200,ZNF513,CRB1,PDE6B,RP9,FAM161A(@>(MARCH9(8.30e@03);KLHL7,SPATA7,CRB1,PDE6B,FAM161A(@>(NETO1(9.25e@03);TTC8,SNRNP200(@>(MYCN(5.63e@03);(@>(TRH(1.81e@04);IDH3B(@>(ADSSL1(4.55e@02);RPGR,RP2,SEMA4A(@>(METTL9(4.57e@02);(@>(CDR2(1.43e@02);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,RP1,PRCD,RPE65,SAG,RBP3,ABCA4(@>(KCNJ13(6.03e@12)1 1Chondrodystrophy GBM 0.47 FGFR3 0.0001457 FGFR3(@>(Syndecan@3@mediated(signaling(events(EGFR,FGFR3);FGFR3(@>(FGFR3b(ligand(binding(and(activation(FGFR3);FGFR3(@>(FGFR3c(ligand(binding(and(activation(FGFR3);FGFR3(@>(FGF(signaling(pathway(AKT1,FGFR3,PIK3CA,PIK3R1,MET)1 1 1Osteogenesis(Imperfecta GBM 1 1 0.078898011 CRTAP,LEPRE1(@>(AKT1(1.95e@02);CRTAP,LEPRE1(@>(CDKN2B(1.15e@02);(@>(MDM2(3.45e@02);COL1A2,COL1A1(@>(KLHL9(1.59e@02);(@>(TRH(3.95e@02);(@>(SMYD3(3.49e@02);(@>(TSPAN31(1.57e@02);CRTAP,LEPRE1(@>(INF2(3.52e@02);LEPRE1(@>(MDK(1.92e@02);COL1A2(@>(WNT2(1.12e@02);(@>(CDK4(2.07e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(CDH13(2.03e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(PTPN21(1.38e@02);CRTAP,LEPRE1(@>(CNIH4(1.64e@02);COL1A2,COL1A1,LEPRE1(@>(CDR2(1.55e@02)1 1Anophthalmos/Micropthalmos GBM 1 1 0.066215584 BMP4,STRA6,SOX2,RAX(@>(MDK(7.73e@03)1 1Long(QT(Syndrome KICH 1 1 0.045450368 SCN5A,CACNA1C(@>(PXDNL(4.73e@03);CACNA1C,CAV3(@>(FBXL22(2.95e@02)1 1Chronic(Granulomatous(Disease KICH 1 1 0.088531104 NCF2,NCF4,CYBB,CYBA(@>(USP3(1.59e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.76e@02);NCF2,NCF4,CYBB,CYBA(@>(RB1CC1(1.86e@02)1 1Polycystic(Kidney,(Autosomal(Dominant KICH 1 1.61E@09 TSC2(@>(mtor(signaling(pathway(PTEN);TSC2(@>(Direct(p53(effectors(PTEN,TP53);TSC2(@>(LKB1(signaling(events(TP53);TSC2(@>(mTOR(signaling(pathway(RB1CC1)1 1 1Spinocerebellar(Ataxia KICH 1 1.66E@10 ATM(@>(apoptotic(signaling(in(response(to(dna(damage(TP53);ATM(@>(regulation(of(cell(cycle(progression(by(plk3(TP53);ATM(@>(p53(pathway(TP53);ATM(@>(cell(cycle:(g2/m(checkpoint(TP53);ATM(@>(Autodegradation(of(the(E3(ubiquitin(ligase(COP1(TP53);ATM(@>(role(of(brca1(brca2(and(atr(in(cancer(susceptibility(TP53);ATM(@>(BARD1(signaling(events(TP53);ATM(@>(atm(signaling(pathway(TP53);ATM(@>(rb(tumor(suppressor/checkpoint(signaling(in(response(to(dna(damage(TP53);ATM(@>(hypoxia(and(p53(in(the(cardiovascular(system(TP53);TBP(@>(Glucocorticoid(receptor(regulatory(network(TP53)0.021413248 JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(SNTG1(1.92e@03);JPH3,ZNF592,CACNA1A,POLG,ATXN7,TDP1,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SETX,PPP2R2B,ATXN1(@>(HERC1(1.92e@03);JPH3,APTX,CACNA1A,ATXN2,SYNE1,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1(@>(ST18(3.88e@03);JPH3,ZNF592,ATXN7,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,ITPR1,SETX,PPP2R2B,ATXN1(@>(RB1CC1(1.37e@02)1 1Combined(Heart(and(Skeletal(Defects KICH 1 0.000738708 EP300,CREBBP(@>(Direct(p53(effectors(PTEN,TP53);EP300,CREBBP(@>(p53(pathway(TP53);EP300(@>(cell(cycle:(g2/m(checkpoint(TP53);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53);EP300(@>(hypoxia(and(p53(in(the(cardiovascular(system(TP53);EP300,CREBBP(@>(Glucocorticoid(receptor(regulatory(network(TP53);CREBBP(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN);EP300,CREBBP(@>(Signaling(events(mediated(by(HDAC(Class(III(TP53)0.448457332 0 CREBBP(@>(TP53,TP53 1Tuberous(Sclerosis KICH 1 1.61E@09 TSC2,TSC1(@>(mtor(signaling(pathway(PTEN);TSC2(@>(Direct(p53(effectors(PTEN,TP53);TSC2,TSC1(@>(LKB1(signaling(events(TP53);TSC2,TSC1(@>(mTOR(signaling(pathway(RB1CC1)0.839245915 1 1Severe(Combined(Immunodeficiency KICH 1 1 0.07356062 ZAP70,IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(USP3(9.65e@03)1 1Specified(Hamartoses KICH 0.13 PTEN 8.15E@35 PTEN(@>(skeletal(muscle(hypertrophy(is(regulated(via(akt@mtor(pathway(PTEN);PTEN(@>(regulation(of(eif@4e(and(p70s6(kinase(PTEN);PTEN(@>(Downstream(TCR(signaling(PTEN);PTEN(@>(mtor(signaling(pathway(PTEN);PTEN(@>(Direct(p53(effectors(PTEN,TP53);STK11(@>(LKB1(signaling(events(TP53);PTEN(@>(BCR(signaling(pathway(PTEN);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(TP53);PTEN(@>(RhoA(signaling(pathway(PTEN);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN);PTEN(@>(Negative(regulation(of(the(PI3K/AKT(network(PTEN);PTEN(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN);PTEN(@>(TCR(signaling(in(na&#xef(ve(CD4+(T(cells(PTEN)1 0.0925 PTEN(@>(TP53,TP530.40083333Polycystic(Kidney,(Autosomal(Dominant KIRC 1 0.008828178 TSC2(@>(mtor(signaling(pathway(PTEN,PIK3CA)0.77024073 1 1Spinocerebellar(Ataxia KIRC 1 1 0.010076664 JPH3,ATXN2,TTBK2,KCNC3,SYT14,PPP2R2B(@>(TSPAN19(2.27e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(EPHA6(6.72e@04);JPH3,CACNA1A,ATXN2,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@>(GPRIN1(1.30e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(CSMD1(3.20e@03);JPH3,ZNF592,CACNA1A,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,ITPR1,FGF14,SETX,SYT14,PPP2R2B,ATXN1(@>(FAM200A(4.51e@03);JPH3,ZNF592,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SETX,PPP2R2B,ATXN1(@>(QKI(3.76e@03);ZNF592,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,ATXN1(@>(PIK3CA(7.15e@03);JPH3,CACNA1A,ATXN2,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,SYT14,PPP2R2B(@>(NEFH(3.84e@03);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2B(@>(PARK2(2.54e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(PTPRN2(3.69e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,A������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 0.63027586Hypopituitarism KIRC 1 0.004667886 GH1(@>(regulation(of(eif@4e(and(p70s6(kinase(PIK3CA,PTEN);GH1(@>(mtor(signaling(pathway(PTEN,PIK3CA);BTK(@>(BCR(signaling(pathway(PIK3CA,PTEN)1 1 1Combined(Heart(and(Skeletal(Defects KIRC 1 0.098019822 EP300(@>(hypoxia@inducible(factor(in(the(cardivascular(system(VHL);EP300,CREBBP(@>(HIF@2@alpha(transcription(factor(network(VHL,TCEB1);EP300,CREBBP(@>(il@7(signal(transduction(PIK3CA);CREBBP(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PIK3CA,PTEN)0.667858981 1 1Tuberous(Sclerosis KIRC 1 0.008738373 TSC2,TSC1(@>(mtor(signaling(pathway(PTEN,PIK3CA)0.455452649 1 1Severe(Combined(Immunodeficiency KIRC 1 1 0.015958615 ZAP70,IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(VHL(1.20e@03);CIITA,RFXAP,DCLRE1C,RAG2,RAG1(@>(PBRM1(2.88e@02)1 1Specified(Hamartoses KIRC 0.03 VHL,PTEN 9.56E@72 PTEN(@>(skeletal(muscle(hypertrophy(is(regulated(via(akt@mtor(pathway(PIK3CA,PTEN);PTEN(@>(regulation(of(eif@4e(and(p70s6(kinase(PIK3CA,PTEN);VHL(@>(vegf(hypoxia(and(angiogenesis(VHL,PIK3CA);PTEN(@>(mtor(signaling(pathway(PTEN,PIK3CA);VHL(@>(hypoxia@inducible(factor(in(the(cardivascular(system(VHL);PTEN(@>(BCR(signaling(pathway(PIK3CA,PTEN);VHL(@>(HIF@2@alpha(transcription(factor(network(VHL,TCEB1);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(VHL,TCEB1,CDKN2A);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN,PIK3CA);PTEN(@>(Negative(regulation(of(the(PI3K/AKT(network(PTEN);PTEN(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PIK3CA,PTEN);PTEN(@>(CXCR4@mediated(signaling(events(PIK3CA,PTEN)0.40794741 0 PTEN(@>(PIK3CA;SDHB(@>(CDKN2A,UQCRFS1,UQCRFS1,TCEB11Chronic(Granulomatous(Disease KIRP 1 1 0.087493235 NCF4(@>(SYTL3(3.78e@02);NCF2,NCF4,CYBB,CYBA(@>(MARCH1(2.06e@02);NCF2,NCF4,CYBB,CYBA(@>(SOD2(3.13e@02);NCF2,NCF4,CYBB,CYBA(@>(KLHL2(2.56e@02);NCF2,CYBB,CYBA(@>(PID1(1.85e@02);NCF2,CYBB,CYBA(@>(SNX9(1.69e@02);NCF2,NCF4,CYBB,CYBA(@>(HRH2(3.13e@02);NCF2,NCF4(@>(CEACAM8(2.87e@02);NCF2,NCF4,CYBB,CYBA(@>(TAGAP(2.62e@02);NCF2,NCF4(@>(FBXL13(2.99e@02);NCF2,NCF4,CYBB(@>(PIK3CB(2.68e@02);NCF2,NCF4,CYBB,CYBA(@>(IGF2R(2.31e@02);NCF2,NCF4,CYBB,CYBA(@>(FNIP2(1.64e@02);NCF2,NCF4,CYBB,CYBA(@>(DOT1L(1.53e@02);NCF2,NCF4,CYBB,CYBA(@>(WTAP(1.85e@02);NCF2,NCF4,CYBB,CYBA(@>(UIMC1(1.95e@02);NCF2,CYBB(@>(NAF1(2.69e@02)1 1Spinocerebellar(Ataxia KIRP 1 1 0.020656922 JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(ZDHHC14(3.03e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(CPE(3.40e@03);JPH3,ZNF592,ATM,TTBK2,KCNC3,ITPR1,SETX,PPP2R2B(@>(MARCH1(3.33e@02);ZNF592,CACNA1A,TTBK2,TBP,KCNC3,ITPR1,FGF14,AFG3L2(@>(PPARGC1B(7.59e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(RAPGEF2(2.75e@03);SPTBN2,KCNC3,SYT14(@>(RXFP1(4.50e@02);JPH3,ZNF592,ATXN7,ATXN2,TDP1,SYNE1,ATM,TTBK2,TBP,KCNC3,ITPR1,FGF14,AFG3L2,PPP2R2B(@>(MAP3K4(4.34e@03);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2B(@>(PARK2(2.75e@03);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B,ATXN1(@>(KIAA0825(2.35e@03);JPH3,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@>(PCDH11X(1.81e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,I����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Neurofibromatosis KIRP 0.47 NF2 1 0.720639739 0 NF1(@>(NF2,CDKN2A 1Severe(Combined(Immunodeficiency KIRP 1 1 0.003470279 ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(SYTL3(6.34e@03);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D(@>(C6orf99(3.25e@03);IL2RG,JAK3,PTPRC,DCLRE1C(@>(HRH2(3.70e@02);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,DCLRE1C(@>(TAGAP(1.66e@02);IL2RG,ADA,PNP,PTPRC(@>(WTAP(3.67e@02);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C(@>(BCL11B(1.22e@03);IL2RG,RFXAP,RFX5,AK2,CD3D,DCLRE1C(@>(VRK1(2.43e@04);IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,DCLRE1C(@>(UIMC1(6.16e@03);AK2,IL7R(@>(RPL22(1.82e@04);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C(@>(IL32(4.78e@04);ADA,PNP(@>(NAF1(2.66e@02);NHEJ1,RFXANK,AK2(@>(TREX2(5.68e@03)1 1Lipoprotein(Deficiencies KIRP 1 1 0.052743895 MTTP,APOB,LCAT,SAR1B,APOA1(@>(ETFDH(6.51e@03);APOB,LCAT,APOA1(@>(LPA(5.65e@03);ABCA1(@>(PID1(3.54e@02);MTTP,APOB,APOA1(@>(CEACAM1(2.64e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(SLC22A1(6.51e@03);(@>(SLC22A3(2.64e@02);MTTP(@>(KCNK5(4.44e@02)1 1Disorders(of(Urea(Cycle(Metabolism KIRP 1 1 0.088531104 ASS1,NAGS,ARG1,ASL,CPS1(@>(ETFDH(1.60e@02);ASS1,ARG1,CPS1(@>(LPA(1.67e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(SLC22A1(1.67e@02);NAGS(@>(MSMO1(1.63e@02)1 1Chronic(Granulomatous(Disease LGG 1 0.438518035 0.076994822 (@>(CRLF2(2.65e@02);CYBA(@>(IRF4(2.99e@02);NCF2,NCF4,CYBB,CYBA(@>(DUSP22(2.36e@02);CYBA(@>(TWF2(1.86e@02);NCF2,NCF4,CYBB,CYBA(@>(ADAM8(1.08e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.20e@02);NCF2,NCF4,CYBB,CYBA(@>(ATP9B(1.09e@02);NCF2,NCF4,CYBB,CYBA(@>(NFATC1(1.29e@02);NCF2,NCF4,CYBB,CYBA(@>(CSF2RA(1.43e@02);NCF2,NCF4,CYBB,CYBA(@>(METRNL(1.91e@02);NCF2,NCF4,CYBB,CYBA(@>(LRRK2(1.06e@02);NCF2,NCF4,CYBA(@>(B3GNTL1(1.22e@02);NCF2,NCF4,CYBB,CYBA(@>(KCNQ1(1.10e@02);NCF2,NCF4,CYBB,CYBA(@>(GLYCTK(1.33e@02);NCF2,NCF4(@>(PHF8(4.49e@02);NCF2,NCF4,CYBA(@>(MEF2D(1.52e@02);NCF2,CYBB,CYBA(@>(KMO(1.10e@02);NCF2,NCF4,CYBB,CYBA(@>(CTDP1(1.08e@02);NCF2,NCF4,CYBB,CYBA(@>(TNFSF13B(4.32e@02);NCF2,NCF4,CYBA(@>(AGAP2(2.75e@02);NCF2,NCF4,CYBB,CYBA(@>(IL3RA(1.77e@02);NCF2,NCF4,CYBB,CYBA(@>(PQLC1(1.26e@02);NCF2,NCF4(@>(IRS2(1.96e@02);NCF2,NCF4,CYBB,CYBA(@>(LPAR6(2.36e@02);NCF4,CYBB,CYBA(@>(STAB1(2.69e@02);NCF2,NCF4,CYBB,CYBA(@>(RB1(3.25e@02);CYBA(@>(HEATR3(3.40e@02);NCF2(@>(P2RY8(1.19e@02);NCF2,NCF4,CYBA(@>(RASA3(1.88e@02);NCF2,NCF4,CYBB,CYBA(���������������������������������������������������������������������������������1 1Congenital(Ichthyosis LGG 1 0.297294297 0.009387698 ALOX12B,ALOXE3,CSTA,TGM1(@>(PARD6G(3.97e@02);(@>(SLC25A6(1.39e@03);CSTA,TGM1(@>(CYP27B1(4.66e@02);ALOX12B,ALDH3A2,ALOXE3,SPINK5,CSTA,KRT2,ABCA12,TGM1(@>(ZNF750(6.06e@04);ALOX12B,ALOXE3,SPINK5,CSTA,ABCA12,TGM1(@>(MPZL2(2.34e@03);ALOX12B,ALDH3A2,ALOXE3,SPINK5,CSTA,ABCA12,TGM1(@>(KLF5(1.39e@03);CSTA,NIPAL4,LIPN,ABHD5(@>(NOTCH1(1.90e@02)1 1Inherited(Adrenogenital(Disorders LGG 1 0.709981324 0.078663977 HSD3B2,POR,CYP17A1,CYP21A2(@>(ITIH1(1.15e@02);HSD3B2,POR,CYP17A1,CYP21A2(@>(ITIH3(2.20e@02);HSD3B2,POR,CYP17A1,CYP21A2(@>(ITIH4(2.20e@02);HSD3B2,CYP17A1,CYP21A2(@>(ARSE(2.53e@02)1 1Pervasive,(Specified(Congenital(Anomalies LGG 1 FGD1 4.83E@30 HRAS,RAF1,SOS1(@>(t(cell(receptor(signaling(pathway(NFATC1,CD3E,PIK3R1,PIK3CA);FGD1(@>(Regulation(of(CDC42(activity(FARP2,FGD1);HRAS,SOS1(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,PIK3R1,PIK3CA);HRAS,NRAS,SOS1,KRAS(@>(Trk(receptor(signaling(mediated(by(PI3K(and(PLC@gamma(PIK3CA,AGAP2,PIK3R1);HRAS,RAF1,SOS1(@>(pdgf(signaling(pathway(PIK3CA,PDGFRA,PIK3R1);PTCH1(@>(Signaling(events(mediated(by(the(Hedgehog(family(PIK3CA,PIK3R1);SOS1(@>(PDGFR@alpha(signaling(pathway(PIK3CA,PDGFRA,PIK3R1);HRAS,SOS1(@>(vegf(hypoxia(and(angiogenesis(PIK3CA,PIK3R1);PTPN11,SOS1(@>(IL2(signaling(events(mediated(by(STAT5(PIK3CA,PIK3R1);PTPN11(@>(VEGFR1(specific(signals(PIK3CA,PIK3R1);HRAS,RAF1(@>(Nongenotropic(Androgen(signaling(PIK3CA,PIK3R1);HRAS,RAF1,SOS1(@>(tpo(signaling(pathway(PIK3CA,PIK3R1);HRAS,RAF1,SOS1(@>(multiple(antiapoptotic(pathways(from(igf@1r(signaling(lead(to(bad(phosphorylation(PIK3CA,PIK3R1);HRAS,RAF1,PTPN11,SOS1(@>(IGF1(pathway(PIK3CA,IRS2,PIK3R1);HRAS,RAF1,SOS1(@>(inhibition(of(cellular(proliferation(by(gleevec(P�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.218652765 0.73112 0.26455Diamond@Blackfan(Anemia LGG 1 1 0.020326331 RPS26,RPS24,RPL5,RPS7(@>(BORA(1.65e@03);RPS24,RPL5,RPS7(@>(GNL3(5.85e@03);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(PTBP1(5.88e@03);RPS26(@>(CDKN2A(1.01e@02);RPS26,RPS19,RPS10(@>(TWF2(3.53e@02);RPL5(@>(TLR9(4.05e@02);(@>(VENTX(8.75e@03);RPS26,RPS19,RPS7(@>(TP53(1.87e@03);(@>(UTF1(9.49e@03);RPS26,RPS7(@>(METTL1(2.67e@02);RPS26(@>(CDK4(2.75e@02);RPS26,RPS19,RPS10,RPS7(@>(SPCS1(6.27e@03);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(TXNL4A(3.42e@02);RPS26,RPS7(@>(HEATR3(9.43e@03);(@>(IL32(2.72e@02);RPS26,RPS7(@>(MZT1(1.39e@02)1 1Inherited(Anomalies(of(the(Skin LGG 1 0.013851661 ATP2A2(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,PIK3R1,PIK3CA);TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(TP53,RB1);TERT(@>(role(of(nicotinic(acetylcholine(receptors(in(the(regulation(of(apoptosis(PIK3CA,PIK3R1)0.112939761 1 1Spinocerebellar(Ataxia LGG 1 1 0.020326331 ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX(@>(PIBF1(2.73e@02);(@>(PPP2R3B(2.73e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(NISCH(4.60e@03);JPH3,ATXN2,SYNE1,SPTBN2,TTBK2,KCNC3,PRKCG,SYT14,PPP2R2B(@>(IFLTD1(1.08e@02);JPH3,ZNF592,CACNA1A,ATXN7,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B,ATXN1(@>(ATRX(1.74e@03);POLG,ATXN7,TDP1,ATM,TTBK2,TBP,ITPR1(@>(MARCH9(9.72e@03);ZNF592,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,ATXN1(@>(PIK3CA(1.35e@02);JPH3,ZNF592,ATXN2,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,PPP2R2B(@>(PIK3C2B(1.06e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(PALM(3.34e@02);APTX,ATXN7,TDP1,ATM,TBP,NOP56,SETX,C10orf2(@>(THOC1(3.47e@02);ZNF592,POLG,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,ATXN1(@>(MDM4(2.02e@02);TDP1,TBP,ITPR1,NOP56,SETX,C10orf2,ATXN1(@>(FUBP1(2.67e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(IGSF9B(1.67e@03);JPH3,APTX,�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Hypopituitarism LGG 1 0.001321061 GH1(@>(regulation(of(eif@4e(and(p70s6(kinase(PIK3CA,PIK3R1,PTEN);GLI2(@>(Signaling(events(mediated(by(the(Hedgehog(family(PIK3CA,PIK3R1);GH1(@>(mtor(signaling(pathway(PTEN,PIK3R1,PIK3CA);BTK(@>(BCR(signaling(pathway(NFATC1,PIK3CA,PIK3R1,PTEN);GH1(@>(growth(hormone(signaling(pathway(PIK3CA,PIK3R1);BTK(@>(EPO(signaling(pathway(PIK3R1,IRS2);GH1(@>(akt(signaling(pathway(PIK3CA,PIK3R1)0.15528125 1 1Combined(Heart(and(Skeletal(Defects LGG 1 0.002464787 CREBBP(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,PIK3R1,PIK3CA);EP300,CREBBP(@>(p53(pathway(TP53,CDKN2A,MDM4);EP300(@>(role(of(mef2d(in(t@cell(apoptosis(NFATC1,CD3E,MEF2D);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53,RB1);EP300,CREBBP(@>(E2F(transcription(factor(network(CDKN2C,CDKN2A,RB1);EP300,CREBBP(@>(il@7(signal(transduction(PIK3CA,PIK3R1);CREBBP(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,PIK3CA,PIK3R1,PIK3C2B);EP300(@>(ATF@2(transcription(factor(network(PDGFRA,CDK4,RB1);EP300,CREBBP(@>(FOXM1(transcription(factor(network(CDKN2A,CDK4,RB1)0.651597416 0.55437288 0.63027586Hereditary(Sensory(Neuropathy LGG 1 1 0.099201187 (@>(DHRSX(1.98e@02);(@>(SLC25A6(4.63e@02);(@>(ASMTL(1.97e@02);LMNA(@>(CD99(4.74e@02);SBF2,PRX,LMNA,HSPB8,MTMR2,SH3TC2,PMP22,KIF1B,NEFL,INF2(@>(SASH1(1.90e@02);(@>(ZBED1(1.98e@02)1 1Severe(Combined(Immunodeficiency LGG 1 2.73E@09 CD3D,PTPRC,ZAP70(@>(t(cell(receptor(signaling(pathway(NFATC1,CD3E,PIK3R1,PIK3CA);CD3D(@>(Downstream(TCR(signaling(PTEN,CD3E);JAK3,IL2RG(@>(IL2(signaling(events(mediated(by(STAT5(PIK3CA,PIK3R1);PTPRC(@>(BCR(signaling(pathway(NFATC1,PIK3CA,PIK3R1,PTEN);CD3D,PTPRC,ZAP70(@>(role(of(mef2d(in(t@cell(apoptosis(NFATC1,CD3E,MEF2D);CD3D(@>(the(co@stimulatory(signal(during(t@cell(activation(PIK3CA,CD3E,PIK3R1);JAK3,IL2RG(@>(il@7(signal(transduction(PIK3CA,PIK3R1);JAK3,IL2RG(@>(IL4@mediated(signaling(events(PIK3CA,PIK3R1,IRS2,IRF4);JAK3,IL2RG(@>(IL2@mediated(signaling(events(PIK3CA,IRS2,PIK3R1);CD3D,PTPRC(@>(CXCR4@mediated(signaling(events(PIK3CA,CD3E,PIK3R1,PTEN)0.001993726 ZAP70,IL2RG,JAK3,RFXAP,IL7R,CD3D,DCLRE1C(@>(PIBF1(2.80e@02);IL2RG,CIITA,RFX5,DCLRE1C(@>(IRF4(1.12e@03);ZAP70,IL2RG,JAK3,RFXAP,RFX5,IL7R,CD3D,DCLRE1C(@>(MARCH9(1.01e@02);IL2RG,RFXAP,RFX5,AK2,CD3D,DCLRE1C(@>(BORA(1.01e@04);ZAP70,IL2RG,RFXAP,RFX5,AK2,IL7R,CD3D,DCLRE1C(@>(THOC1(6.29e@03);ZAP70,IL2RG,CIITA,JAK3,RFXAP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1(@>(MDM4(3.01e@03);IL2RG,JAK3,PNP,RFX5,PTPRC(@>(ADAM8(4.10e@02);CIITA,RFX5,DCLRE1C(@>(TLR9(8.90e@05);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1(@>(CD3E(3.13e@04);ZAP70,IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(ATP9B(1.20e@02);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC(@>(NFATC1(1.78e@02);IL2RG,RFXAP,AK2,DCLRE1C(@>(FUBP1(1.00e@02);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C(@>(PPM1M(4.94e@03);IL2RG,JAK3,PTPRC,AK2,DCLRE1C(@>(B3GNTL1(1.63e@04);RFXAP,RFX5,AK2,DCLRE1C(@>(NCAPD3(2.56e@03);ZAP70,IL2RG,JAK3,PNP,PTPRC,IL7R,CD3D,DCLRE1C(@>(GLYCTK(1.19e@02);ZAP70,IL2RG,JAK3,RFXAP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(PHF8(1.18e@02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.28232609 1Specified(Hamartoses LGG 0.66 PTEN 1.74E@07 PTEN(@>(skeletal(muscle(hypertrophy(is(regulated(via(akt@mtor(pathway(PIK3CA,PIK3R1,PTEN);PTEN(@>(regulation(of(eif@4e(and(p70s6(kinase(PIK3CA,PIK3R1,PTEN);VHL(@>(vegf(hypoxia(and(angiogenesis(PIK3CA,PIK3R1);PTEN(@>(Downstream(TCR(signaling(PTEN,CD3E);PTEN(@>(mtor(signaling(pathway(PTEN,PIK3R1,PIK3CA);PTEN(@>(BCR(signaling(pathway(NFATC1,PIK3CA,PIK3R1,PTEN);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN,PIK3R1,PIK3CA);PTEN(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PTEN,PIK3CA,PIK3R1,PIK3C2B);PTEN(@>(CXCR4@mediated(signaling(events(PIK3CA,CD3E,PIK3R1,PTEN)0.613735058 0.20252632 PTEN(@>(PIK3CA;STK11(@>(CDKN2A1Holoprosencephaly LGG 1 1 0.043065539 TDGF1,NODAL,FOXH1,GLI2(@>(VENTX(4.40e@03);TDGF1,NODAL,FOXH1,GLI2(@>(UTF1(1.43e@02);TDGF1,FOXH1,ZIC2(@>(PLCXD1(1.13e@02);(@>(MID1(1.43e@02)1 1Li(Fraumeni(and(Related(Syndromes LGG 0.03 CDKN2A,TP53 4.44E@10 TP53(@>(chaperones(modulate(interferon(signaling(pathway(TP53,RB1);TP53,CDKN2A,CHEK2(@>(p53(pathway(TP53,CDKN2A,MDM4);TP53(@>(estrogen(responsive(protein(efp(controls(cell(cycle(and(breast(tumors(growth(CDK4,TP53);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(TP53,RB1);TP53(@>(btg(family(proteins(and(cell(cycle(regulation(TP53,RB1);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53(@>(p53(signaling(pathway(CDK4,TP53,RB1);TP53(@>(regulation(of(transcriptional(activity(by(pml(TP53,RB1);TP53(@>(rb(tumor(suppressor/checkpoint(signaling(in(response(to(dna(damage(CDK4,TP53,RB1);CDKN2A(@>(E2F(transcription(factor(network(CDKN2C,CDKN2A,RB1);TP53(@>(p75(NTR)@mediated(signaling(PIK3CA,TP53,PIK3R1);CDKN2A,CHEK2(@>(FOXM1(transcription(factor(network(CDKN2A,CDK4,RB1)0.667858981 0 CDKN2A(@>(CDKN2C;CHEK2(@>(CDK4;TP53(@>(PTEN,TP53,GNL30.31565625Lipoprotein(Deficiencies LGG 1 0.563318142 0.041865488 APOB,SAR1B(@>(IDH1(1.07e@02);MTTP,APOB,LCAT,APOA1,ABCA1(@>(GLYCTK(8.69e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ITIH1(7.83e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ITIH3(7.83e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ITIH4(4.19e@03);MTTP,APOB,LCAT,APOA1(@>(ARSD(6.80e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ARSE(6.11e@03)1 1Disorders(of(Urea(Cycle(Metabolism LGG 1 0.01382358 ARG1(@>(IL4@mediated(signaling(events(PIK3CA,PIK3R1,IRS2,IRF4);ARG1(@>(ATF@2(transcription(factor(network(PDGFRA,CDK4,RB1)0.066215584 ASS1,NAGS,ASL,CPS1(@>(HSBP1L1(7.80e@03);NAGS,ARG1,ASL,CPS1(@>(GLYCTK(3.87e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ITIH1(2.55e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ITIH3(1.30e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ITIH4(2.55e@02);NAGS,ARG1,ASL(@>(IDH2(2.02e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ARSD(8.13e@03);ASS1,NAGS,ARG1,ASL,CPS1(@>(ARSE(8.74e@03)1 1Retinitis(Pigmentosa LGG 1 0.611253215 1.45E@10 SNRNP200,ZNF513,CRB1,PDE6B,RP9,FAM161A(@>(MARCH9(6.27e@03);(@>(DHRSX(2.24e@03);CRX,FSCN2,RDH12,PRPH2,CNGB1,EYS,CRB1,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3(@>(ASMT(5.31e@09);TTC8,KLHL7,EYS,SPATA7,CRB1,MERTK,PDE6B,CERKL,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(PCDHAC2(9.96e@03);IMPDH1(@>(CYP27B1(1.88e@02);(@>(ZBED1(5.82e@03);CRX,FSCN2,RDH12,SNRNP200,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,USH2A,PROM1,CNGA1,RP1,PRCD,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(GLB1L2(2.11e@12);CRX,FSCN2,PRPH2,CNGB1,EYS,CRB1,CERKL,TULP1,PRCD,IMPG2,C2orf71,RBP3(@>(MYO16(2.31e@05);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,RP1,PRCD,RPE65,SAG,RBP3,ABCA4(@>(KCNJ13(4.64e@12);FAM161A(@>(SHOX(3.26e@10);SPATA7,CRB1,PDE6B(@>(DRD1(1.42e@02)0.57574242 1Osteogenesis(Imperfecta LGG 1 0.521588572 0.076277809 (@>(SLC25A6(2.01e@02);CRTAP,LEPRE1(@>(TWF2(1.35e@02);CRTAP,COL1A2(@>(METRNL(2.47e@02);(@>(DCP1B(2.51e@02);COL1A2,COL1A1,LEPRE1(@>(CDC16(1.25e@02);COL1A2,LEPRE1(@>(JAM3(4.12e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(CD99(4.89e@02);(@>(TSPAN31(1.60e@02);CRTAP,COL1A2(@>(NAB2(1.38e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(PDGFRA(3.31e@02);CRTAP(@>(SMIM4(2.47e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(SASH1(1.40e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(FGD1(1.32e@02);COL1A2(@>(XG(2.25e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(NT5DC2(1.17e@02);(@>(CDK4(2.15e@02);LEPRE1(@>(NOX4(1.21e@02);CRTAP,LEPRE1(@>(SPCS1(1.16e@02);CRTAP,COL1A2,COL1A1,LEPRE1(@>(TEAD3(1.91e@02);CRTAP,LEPRE1(@>(TXNL4A(4.07e@02);(@>(SHOX(1.17e@02);COL1A2,COL1A1,LEPRE1(@>(MXRA5(2.58e@02);CRTAP,COL1A2,COL1A1(@>(PRCP(1.20e@02)1 1Anophthalmos/Micropthalmos LGG 1 1 0.070668747 VSX2,MFRP,RAX(@>(GLB1L2(8.83e@03)1 1Chronic(Granulomatous(Disease LUAD 1 0.384190056 0.078679242 NCF4,CYBA(@>(CCND3(3.54e@02);NCF2,NCF4,CYBA(@>(ZGPAT(1.74e@02);NCF2,NCF4,CYBB,CYBA(@>(RIT1(1.48e@02);NCF2,NCF4,CYBB,CYBA(@>(ITGAX(1.79e@02);NCF2,NCF4,CYBB,CYBA(@>(DNAJC5(1.82e@02);NCF2,NCF4,CYBB,CYBA(@>(PTGER4(1.60e@02);NCF2,NCF4,CYBB,CYBA(@>(SIRPB1(1.30e@02);NCF2,NCF4,CYBB,CYBA(@>(TNFSF13B(4.88e@02);NCF2,NCF4,CYBB,CYBA(@>(PQLC1(1.33e@02);NCF2,NCF4,CYBB,CYBA(@>(TBL1X(1.72e@02);NCF4,CYBA(@>(ARHGEF6(2.95e@02);NCF2,NCF4,CYBB,CYBA(@>(GNG2(1.37e@02);NCF2,NCF4,CYBB,CYBA(@>(NFATC1(1.81e@02);NCF2,NCF4(@>(U2AF1(3.61e@02);NCF2,NCF4,CYBB,CYBA(@>(BTK(1.94e@02);NCF2,NCF4,CYBB,CYBA(@>(SAMSN1(1.35e@02);NCF2,NCF4,CYBB,CYBA(@>(LPAR6(2.19e@02);NCF2,NCF4(@>(IL18RAP(1.83e@02);NCF2,NCF4,CYBB,CYBA(@>(RB1(3.72e@02);NCF2,NCF4,CYBB(@>(ADNP2(3.36e@02);NCF2,NCF4,CYBB,CYBA(@>(MFSD7(3.07e@02);NCF2,NCF4,CYBB,CYBA(@>(PMAIP1(1.26e@02);NCF4,CYBB,CYBA(@>(TBX21(1.81e@02);NCF2,NCF4,CYBB,CYBA(@>(ANKRD44(1.82e@02);NCF2,NCF4,CYBB,CYBA(@>(AQP9(1.28e@02);NCF2,NCF4,CYBB,CYBA(@>(PPM1F(1.53e@02);NCF2,NCF4,CYBB,CYBA(@>(CTDP1(1.37e@02);NCF2,NCF4,CYBB,CYB���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Congenital(Ichthyosis LUAD 1 1 0.020062495 ALOX12B,SPINK5,CSTA,TGM1(@>(SPRR3(8.05e@03);ALOX12B(@>(KRT28(3.04e@03);(@>(FLG(1.68e@03);ALOX12B,SPINK5,CSTA,ABCA12,TGM1(@>(SERPINB13(1.68e@03);ALOX12B,SPINK5,KRT2,ABCA12(@>(POF1B(8.05e@03)1 1Disorders(of(Phosphorous(Metabolism LUAD 1 CYP27B1 0.003900738 FGF23(@>(Syndecan@2@mediated(signaling(events(HRAS,LAMA3,FGF19,NF1);CYP27B1(@>(Vitamin(D((calciferol)(metabolism(GC,CYP27B1);FGF23(@>(Syndecan@1@mediated(signaling(events(COL11A1,COL5A2,MET,FGF19,COL5A1,COL3A1)1 1 1Diamond@Blackfan(Anemia LUAD 1 1 0.006070507 RPS26,RPS19,RPS10,RPS7(@>(BYSL(4.40e@02);RPL5,RPS10,RPL11(@>(CCND3(3.53e@02);(@>(UTF1(8.74e@03);RPS24,RPL5(@>(ALG10(3.42e@02);RPS26,RPS19,RPS7(@>(GEMIN4(2.58e@02);RPS26(@>(CDKN2A(9.01e@03);RPS26,RPL35A,RPS7(@>(U2AF1(4.63e@04);RPS26,RPS7(@>(FANCD2(1.47e@02);RPS26,RPS19,RPS10,RPL35A(@>(LAGE3(4.10e@02);(@>(CTCFL(1.14e@02);(@>(MDM2(4.11e@02);(@>(VENTX(7.84e@03);RPS26,RPS19,RPS7(@>(TP53(1.67e@03);RPS26,RPS19,RPS10,RPS7(@>(EIF4EBP1(1.12e@02);RPS26,RPS19,RPS7(@>(TFDP1(2.13e@03);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(TXNL4A(3.14e@02);(@>(TERT(1.10e@02);RPS19,RPS7(@>(CNIH1(4.86e@04);RPL5,RPS7(@>(NRAS(1.31e@03);RPS26,RPS10,RPL35A(@>(RNMTL1(3.54e@04);RPS26,RPS7(@>(METTL1(2.41e@02);RPS26,RPS19,RPS10,RPS7(@>(HAX1(1.49e@02)1 1Inherited(Anomalies(of(the(Skin LUAD 1 TERT 3.07E@07 TERT(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(TERT,TP53);TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(TERT,TP53,RB1,KRAS);KRT1,TERT(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDKN2A,TBL1X,CCND1,APC,TERT,SMARCA4)0.086097896 KRT6A,KRT16(@>(CYP27B1(1.45e@02);WRAP53,TERC,DKC1,NHP2,NOP10(@>(GEMIN4(4.63e@02);TERC,NHP2,NOP10(@>(FAM58A(4.53e@02);KRT6A,TERC,NHP2,KRT16,NOP10(@>(HRAS(4.57e@02);TERC,NHP2,NOP10(@>(SLC10A3(4.68e@02);KRT6A,NHP2,KRT16,NOP10(@>(NXN(4.37e@02);(@>(AKR1B10(4.64e@02)1 1Spinocerebellar(Ataxia LUAD 1 ATM 8.60E@07 PRKCG(@>(EGFR(Inhibitor(Pathway,(Pharmacodynamics(ERBB2,NRAS,HRAS,EGFR,KRAS);ATM(@>(apoptotic(signaling(in(response(to(dna(damage(ATM,TP53);ATM(@>(p53(pathway(CDKN2A,ATM,TP53,MDM2);ATM(@>(ATM(pathway(MDM2,ATM,FANCD2);ATM(@>(cell(cycle:(g2/m(checkpoint(MDM2,ATM,TP53,MYT1);ATM(@>(ATM(mediated(response(to(DNA(double@strand(break(ATM);ATM(@>(cdc25(and(chk1(regulatory(pathway(in(response(to(dna(damage(ATM,MYT1);ATM(@>(role(of(brca1(brca2(and(atr(in(cancer(susceptibility(ATM,TP53,FANCD2);ATM(@>(BARD1(signaling(events(ATM,TP53,FANCD2);ATM(@>(atm(signaling(pathway(MDM2,ATM,TP53);PRKCG(@>(IL8@(and(CXCR2@mediated(signaling(events(PLCB1,GNG2,HCK);ATM(@>(rb(tumor(suppressor/checkpoint(signaling(in(response(to(dna(damage(ATM,TP53,RB1,MYT1);PRKCG(@>(IL8@(and(CXCR1@mediated(signaling(events(PLCB1,GNG2,HCK);ATM(@>(Metformin(Pathway,(Pharmacodynamic(ATM,STK11,PRKAA1);ATM(@>(E2F(transcription(factor(network(TFDP1,CCND3,ATM,CDKN2A,RB1);ATM(@>(hypoxia(and(p53(in(the(cardiovascular(system(MDM2,ATM,TP53);TBP(@>(Glucocorticoid(rece����������������������������������������������������������������������������������������������������������������������������������������������������0.001274155 SYT14(@>(SCG2(2.09e@02);APTX,ZNF592,ATXN2,TTBK2,TBP,KCNC3,ITPR1,NOP56,SETX,SYT14,C10orf2(@>(KIAA0907(2.58e@03);JPH3,TDP1,KCNC3,FGF14(@>(SAMD10(1.63e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(DOC2B(3.57e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(C1orf173(3.57e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(TMEM132D(4.61e@03);JPH3,APTX,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(NF1(2.40e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(OPCML(2.50e@03);JPH3,ZNF592,SPTBN2,TTBK2,KCNC3,PRNP,PRKCG,FGF14,PPP2R2B,ATXN1(@>(DNAJC5(3.21e@02);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B(@>(TTC33(2.02e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Glucose@6@Phosphate(Dehydrogenase(DeficiencyLUAD 0.05 UBL4A,G6PD 0.034828736 NA(@>(NA(NA) @1 1 1Hypopituitarism LUAD 1 BTK 0.034828736 FGFR1(@>(Syndecan@2@mediated(signaling(events(HRAS,LAMA3,FGF19,NF1);BTK(@>(bcr(signaling(pathway(NFATC1,HRAS,BTK,PPP3CA);BTK(@>(EPO(signaling(pathway(HRAS,PTPN11,BTK);POU1F1(@>(Glucocorticoid(receptor(regulatory(network(NFATC1,TP53,TBX21,MDM2,SMARCA4,GATA3);GH1(@>(trefoil(factors(initiate(mucosal(healing(ERBB2,HRAS,EGFR);FGFR1(@>(Syndecan@1@mediated(signaling(events(COL11A1,COL5A2,MET,FGF19,COL5A1,COL3A1)1 0.2886 1Combined(Heart(and(Skeletal(Defects LUAD 1 3.78E@08 EP300,CREBBP(@>(Direct(p53(effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);EP300,CREBBP(@>(p53(pathway(CDKN2A,ATM,TP53,MDM2);EP300(@>(cell(cycle:(g2/m(checkpoint(MDM2,ATM,TP53,MYT1);EP300(@>(Validated(transcriptional(targets(of(AP1(family(members(Fra1(and(Fra2(CCND1,NFATC1,LAMA3,CDKN2A);EP300(@>(Validated(transcriptional(targets(of(TAp63(isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);EP300(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDKN2A,TBL1X,CCND1,APC,TERT,SMARCA4);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53,RB1);EP300,CREBBP(@>(E2F(transcription(factor(network(TFDP1,CCND3,ATM,CDKN2A,RB1);EP300(@>(p73(transcription(factor(network(JAG2,MDM2,NTRK1,RB1,WWOX);EP300(@>(hypoxia(and(p53(in(the(cardiovascular(system(MDM2,ATM,TP53);EP300,CREBBP(@>(Glucocorticoid(receptor(regulatory(network(NFATC1,TP53,TBX21,MDM2,SMARCA4,GATA3);EP300(@>(melanocyte(development(and(pigmentation(pathway(HRAS,KIT)0.71670305 1 1Neurofibromatosis LUAD 0.66 NF1 0.001098828 NF1(@>(Regulation(of(Ras(family(activation(HRAS,NRAS,NF1,KRAS);NF1(@>(Syndecan@2@mediated(signaling(events(HRAS,LAMA3,FGF19,NF1);NF1(@>(chromatin(remodeling(by(hswi/snf(atp@dependent(complexes(ARID1A,SMARCA4,NF1)0.675271975 0.22904762 NF2(@>(CDKN2A 0.60125Hereditary(Sensory(Neuropathy LUAD 1 NTRK1 1.95E@05 NTRK1(@>(Trk(receptor(signaling(mediated(by(PI3K(and(PLC@gamma(CCND1,HRAS,NRAS,NTRK1,KRAS);NTRK1(@>(ARMS@mediated(activation(NTRK1,BRAF);NDRG1(@>(Direct(p53(effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);EGR2(@>(Validated(transcriptional(targets(of(TAp63(isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);NTRK1(@>(TRKA(activation(by(NGF(NTRK1);NTRK1(@>(NGF(signalling(via(TRKA(from(the(plasma(membrane(NTRK1);NTRK1(@>(Signalling(to(ERKs(NTRK1);NTRK1(@>(Signalling(to(STAT3(NTRK1);NTRK1(@>(trka(receptor(signaling(pathway(HRAS,NTRK1);RAB7A(@>(IL8@(and(CXCR2@mediated(signaling(events(PLCB1,GNG2,HCK);NTRK1(@>(Frs2@mediated(activation(NTRK1,BRAF);NTRK1(@>(Signalling(to(p38(via(RIT(and(RIN(NTRK1,BRAF);NTRK1(@>(p73(transcription(factor(network(JAG2,MDM2,NTRK1,RB1,WWOX);PMP22(@>(a6b1(and(a6b4(Integrin(signaling(ERBB2,MET,HRAS,LAMA3,EGFR)0.128961218 1 1Severe(Combined(Immunodeficiency LUAD 1 0.009297242 DCLRE1C(@>(ATM(pathway(MDM2,ATM,FANCD2);ADA(@>(Validated(transcriptional(targets(of(TAp63(isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);ADA(@>(p73(transcription(factor(network(JAG2,MDM2,NTRK1,RB1,WWOX);JAK3,IL2RG(@>(IL2@mediated(signaling(events(HRAS,NRAS,PTPN11,KRAS);JAK3(@>(il(6(signaling(pathway(PTPN11,HRAS);ADA(@>(Validated(transcriptional(targets(of(deltaNp63(isoforms(COL5A1,CDKN2A,ATM,MDM2)0.001684618 CIITA,RFX5,DCLRE1C(@>(C11orf35(2.10e@02);(@>(GATA3(2.09e@02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1(@>(CCND3(1.40e@04);ZAP70,IL2RG,JAK3,ADA,PTPRC,IL7R,CD3D(@>(ZGPAT(2.99e@03);ZAP70,IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D(@>(PTGER4(1.27e@02);IL2RG,CIITA,JAK3,ADA,PNP,PTPRC(@>(TNFSF13B(4.44e@02);RFXANK(@>(RBM10(2.42e@03);ZAP70,IL2RG,RFXAP,RFX5,AK2,IL7R,CD3D,DCLRE1C(@>(ALG10(7.50e@04);IL2RG,RFX5,DCLRE1C(@>(CMTR2(2.45e@03);IL2RG,JAK3,PNP,RFX5,PTPRC,DCLRE1C(@>(AOAH(2.05e@02);IL2RG,JAK3,PTPRC(@>(TBL1X(2.39e@02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(ARHGEF6(9.58e@04);ZAP70,IL2RG,CIITA,JAK3,RFXAP,RFX5,PTPRC,CD3D,DCLRE1C(@>(ARID1A(5.42e@03);ZAP70,IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,DCLRE1C(@>(GNG2(3.39e@02);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC(@>(NFATC1(1.94e@02);ADA,PNP,AK2(@>(U2AF1(9.03e@04);ZAP70,IL2RG,JAK3,RFXAP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1(@>(ARID2(1.03e@02);IL2RG,CIITA,JAK3,RFX5,PTPRC,DCLRE1C(@>(BTK(3.08e@03);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1(@>(THEMIS(2.35e@04);RFXA��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Specified(Hamartoses LUAD 0.97 STK11 2.59E@05 VHL(@>(vegf(hypoxia(and(angiogenesis(HRAS,KDR,ARNT);PTEN(@>(Direct(p53(effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53,ARNT);STK11(@>(Metformin(Pathway,(Pharmacodynamic(ATM,STK11,PRKAA1)0.717635623 0.25727907 1Li(Fraumeni(and(Related(Syndromes LUAD 0.09 CDKN2A,TP53 4.67E@28 TP53(@>(chaperones(modulate(interferon(signaling(pathway(TP53,RB1);TP53(@>(apoptotic(signaling(in(response(to(dna(damage(ATM,TP53);TP53(@>(Direct(p53(effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);TP53,CDKN2A,CHEK2(@>(p53(pathway(CDKN2A,ATM,TP53,MDM2);CHEK2(@>(ATM(pathway(MDM2,ATM,FANCD2);TP53,CHEK2(@>(cell(cycle:(g2/m(checkpoint(MDM2,ATM,TP53,MYT1);CDKN2A(@>(Validated(transcriptional(targets(of(AP1(family(members(Fra1(and(Fra2(CCND1,NFATC1,LAMA3,CDKN2A);CDKN2A(@>(Validated(transcriptional(targets(of(TAp63(isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);TP53(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(TERT,TP53);TP53,CHEK2(@>(role(of(brca1(brca2(and(atr(in(cancer(susceptibility(ATM,TP53,FANCD2);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53,ARNT);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(TERT,TP53,RB1,KRAS);CDKN2A(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDKN2A,TBL1X,CCND1�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.719639888 0 CDKN2A(@>(MDM2;CHEK2(@>(CCND3,ATM;TP53(@>(TP53,SMAD41Genetic(Anomalies(of(Leukocytes LUAD 1 0.02688935 ITGB2(@>(Beta2(integrin(cell(surface(interactions(ITGAX,SPON2,FGB)0.471117151 1 1Lipoprotein(Deficiencies LUAD 1 MTTP 1 0.097427692 APOB,LCAT,SAR1B,APOA1(@>(SLC26A1(2.83e@02);APOB,LCAT,SAR1B,APOA1(@>(PCK1(3.92e@02);APOB,LCAT,SAR1B,APOA1(@>(PROS1(1.82e@02);(@>(AKR1C2(2.10e@02);APOB,LCAT,SAR1B,APOA1(@>(EHHADH(1.83e@02);APOB,LCAT,SAR1B,APOA1(@>(BHMT(1.95e@02);APOB,LCAT,SAR1B,APOA1(@>(GBA3(1.93e@02);APOB,LCAT,SAR1B,APOA1(@>(ABCG5(2.40e@02);APOB,LCAT,SAR1B,APOA1(@>(MTTP(2.00e@02);(@>(CD5L(2.00e@02)1 1Disorders(of(Urea(Cycle(Metabolism LUAD 1 1 0.069332922 ASS1,NAGS,ARG1,ASL,CPS1(@>(GC(1.50e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(SLC26A1(1.99e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(FGB(2.95e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(PROS1(1.02e@02);ASS1,NAGS,ASL,CPS1(@>(HSBP1L1(9.88e@03);ASS1(@>(AKR1C2(3.55e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(EHHADH(1.20e@02);NAGS,ARG1,ASL,CPS1(@>(SOWAHB(1.38e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(CYP4V2(3.55e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(GBA3(8.52e@03);ASS1,NAGS,ARG1,ASL,CPS1(@>(ABCG5(9.51e@03);ASS1,NAGS,ARG1,ASL,CPS1(@>(MTTP(9.88e@03);(@>(CD5L(9.22e@03)1 1Retinitis(Pigmentosa LUAD 1 PDE6B 0.828420314 6.10E@15 RPGR,CRX,SNRNP200,CA4,EYS,CRB1,CERKL,PRPF3,TULP1,C2orf71,TOPORS,FAM161A(@>(KIAA0907(3.68e@02);KLHL7,SPATA7,CRB1,MERTK,CERKL,FAM161A(@>(DOC2B(3.41e@02);(@>(MUC16(2.78e@04);SNRNP200,CRB1,PRPF31,CERKL,IDH3B,PRPF8,FAM161A(@>(UCKL1(2.83e@03);CRX,FSCN2,RDH12,PRPH2,CNGB1,EYS,CRB1,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3(@>(GPR112(5.36e@07);KLHL7,SPATA7,CRB1,PRCD,IMPG2,C2orf71,FAM161A(@>(TMEM132D(3.43e@02);SNRNP200(@>(FZD10(1.68e@02);IMPDH1(@>(CYP27B1(1.33e@02);CRX,SNRNP200,KLHL7,EYS,SPATA7,CRB1,CERKL,USH2A,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(GTF2I(4.98e@02);(@>(SLC22A6(1.68e@02);CNGA1(@>(ANKRD37(1.67e@02);CRX,FSCN2,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,MERTK,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(RIMS2(3.67e@04);SPATA7,CRB1,PRPF3(@>(ITGB8(1.07e@02);RPGR,CA4,IMPDH1,PRPF3,BEST1,RP2,TOPORS,SEMA4A(@>(TBL1X(2.76e@02);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,EYS,CRB1,NRL,PDE6A,PDE6G,CERKL,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,IMPG2,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(RP1L1(1.27e@17);CRX,FSCN�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.02531579 CNGA1(@>(CNGA2,LRRC32,EIF4G3;CNGB1(@>(PABPC3,PDE6B,PDE6B,TXNL4A;IMPDH1(@>(PRPF6,PABPC3;PDE6A(@>(TXNL4A;PDE6G(@>(PRPF6,EIF4G3;PRPF3(@>(PABPC3;PRPF31(@>(TXNL4A;PRPF8(@>(PRPF6,U2AF1,U2AF1;RP9(@>(PABPC3;SNRNP200(@>(TXNL4A1Haemophilia LUAD 0.7 F8 0.071266219 VWF(@>(Platelet(Aggregation(Inhibitor(Pathway,(Pharmacodynamics(COL3A1,COL4A2,FGB,PLCB1);F8,F9(@>(intrinsic(prothrombin(activation(pathway(F8,PROS1,COL4A2,FGB)1 0.7955 0.52680952Chronic(Granulomatous(Disease LUSC 1 1 0.067971546 NCF2,NCF4,CYBB,CYBA(@>(B2M(1.31e@02);NCF2,NCF4,CYBB,CYBA(@>(MARCH1(8.59e@03);NCF2,NCF4,CYBB(@>(USP25(1.12e@02);NCF2,NCF4,CYBB,CYBA(@>(NFE2L2(8.87e@03);NCF2,NCF4,CYBB,CYBA(@>(LYZ(2.21e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.06e@02);NCF2,NCF4,CYBB,CYBA(@>(HDAC10(1.05e@02);NCF2,NCF4,CYBB,CYBA(@>(KDM5A(1.96e@02);NCF2,NCF4,CYBB,CYBA(@>(NFATC1(1.20e@02);NCF4,CYBB,CYBA(@>(CHKB(2.32e@02);NCF2,NCF4,CYBB,CYBA(@>(TRABD(1.01e@02);NCF2,NCF4,CYBB,CYBA(@>(CREBBP(1.06e@02);NCF2,NCF4,CYBB,CYBA(@>(ODF3B(1.41e@02);NCF2,NCF4,CYBB,CYBA(@>(CTDP1(9.25e@03);NCF2,NCF4,CYBB,CYBA(@>(REL(9.24e@03);NCF2,NCF4,CYBB,CYBA(@>(KDM6A(1.20e@02);NCF2,NCF4,CYBB,CYBA(@>(METRNL(9.24e@03);NCF2,NCF4,CYBB,CYBA(@>(PQLC1(8.59e@03);NCF2,NCF4,CYBB,CYBA(@>(LPAR6(1.21e@02);NCF2,NCF4,CYBB,CYBA(@>(BID(1.21e@02);NCF2,NCF4,CYBB,CYBA(@>(RB1(1.92e@02);NCF2,NCF4,CYBB,CYBA(@>(PIM3(8.92e@03);NCF4(@>(NINJ2(1.01e@02);NCF2,NCF4,CYBB,CYBA(@>(EVI2A(1.71e@02);NCF2,NCF4,CYBB,CYBA(@>(EVI2B(9.25e@03);NCF2,NCF4(@>(EXOC3(2.80e@02);NCF2,NCF4,CYBB,CYBA(@>(NOTCH1(8.21e@03);NCF2,NCF4,CY�����������������������������������������������������������1 1Congenital(Ichthyosis LUSC 1 1 0.006200982 ALOX12B,ALOXE3,SPINK5,CSTA,KRT2,ABCA12,TGM1(@>(CERS3(3.87e@04);CSTA,NIPAL4,LIPN,ABHD5(@>(NFE2L2(3.14e@02);ALOX12B,SPINK5,CSTA,TGM1(@>(PAX9(3.14e@02);ALOX12B,ALOXE3,CSTA,TGM1(@>(PARD6G(4.70e@02);CSTA,NIPAL4,LIPN,ABHD5(@>(NOTCH1(2.92e@02);ABCA12,TGM1(@>(EGFR(3.98e@02)0.79744737 1Diamond@Blackfan(Anemia LUSC 1 0.959794083 0.002784462 RPS26(@>(CDKN2A(9.26e@03);(@>(YEATS4(1.30e@03);RPS26,RPS19,RPS7(@>(TP53(1.51e@03);RPS26,RPS19,RPL35A,RPS7(@>(PDCD6(1.33e@04);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(TXNL4A(3.08e@02);(@>(NMU(2.86e@02);RPL5,RPS7(@>(TTF2(4.93e@02);RPS26,RPS19,RPS7(@>(TYMS(2.02e@03);RPS26,RPS7(@>(TRIP13(9.59e@03)1 1Spinocerebellar(Ataxia LUSC 1 0.977629639 0.01980021 JPH3,CACNA1A,POLG,ATXN2,SYNE1,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,PPP2R2B(@>(SBF1(9.84e@03);JPH3,ZNF592,ATM,TTBK2,KCNC3,ITPR1,SETX,PPP2R2B(@>(MARCH1(3.18e@02);JPH3,CACNA1A,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@>(L1CAM(1.18e@02);JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1(@>(CCSER1(2.45e@03);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2B(@>(PARK2(2.80e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(MAPK8IP2(2.90e@03);TDP1,ATM,TBP,NOP56,SYT14,C10orf2(@>(CCDC77(1.54e@02);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(CLCN4(2.99e@03);POLG,ATXN2,TBP,NOP56,AFG3L2,C10orf2(@>(BRD9(2.95e@03);ZNF592,POLG,ATXN7,ATXN2,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,ATXN1(@>(BRD1(3.18e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(CSMD3(2.99e@03);JPH3,CACNA1A��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 0.73607576Combined(Heart(and(Skeletal(Defects LUSC 0.6 CREBBP 3.46E@14 CREBBP(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,CREBBP,PIK3CA);CREBBP(@>(inhibition(of(huntingtons(disease(neurodegeneration(by(histone(deacetylase(inhibitors(CREBBP);EP300,CREBBP(@>(Direct(p53(effectors(TP53,BID,EGFR,RB1,PTEN,CREBBP);EP300,CREBBP(@>(p53(pathway(CDKN2A,TP53,CREBBP);CREBBP(@>(Notch@HLH(transcription(pathway(CREBBP);EP300,CREBBP(@>(mechanism(of(gene(regulation(by(peroxisome(proliferators(via(ppara(FAT1,RB1,CREBBP);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53,CREBBP,RB1);EP300,CREBBP(@>(E2F(transcription(factor(network(TYMS,CDKN2A,CREBBP,RB1);EP300,CREBBP(@>(il@7(signal(transduction(PIK3CA,CREBBP);EP300(@>(p73(transcription(factor(network(MAPK11,CDK6,RB1,WWOX);EP300,CREBBP(@>(Glucocorticoid(receptor(regulatory(network(NFATC1,TP53,MAPK11,CREBBP);CREBBP(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PIK3CA,CREBBP,PTEN);EP300(@>(ATF@2(transcription(factor(network(MAPK11,RB1,NF1);CREBBP(@>(Signaling(events(mediated(by(TCPTP(PIK3CA,EGFR,CREBBP);EP300,CREBBP(@>(F����������������������������������������������������1 0.0925 CREBBP(@>(TP53,TP53;EP300(@>(CREBBP1Neurofibromatosis LUSC 0.53 NF1 0.236599603 0.719411434 0 NF1(@>(EVI2A,CDKN2A;NF2(@>(NF11Severe(Combined(Immunodeficiency LUSC 1 0.23074305 0.001138137 ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(B2M(6.47e@03);IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,DCLRE1C(@>(PLEKHO1(2.51e@02);ZAP70,IL2RG,CIITA,JAK3,PTPRC,IL7R,CD3D,DCLRE1C(@>(BRD1(1.02e@02);(@>(YEATS4(1.46e@02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(HDAC10(2.23e@04);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(KDM5A(8.24e@03);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC(@>(NFATC1(1.79e@02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(FOXP1(2.39e@02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(CHKB(1.16e@03);ZAP70,IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(TRABD(1.29e@04);RFXAP,RFX5,DCLRE1C(@>(PRDM15(8.47e@03);ZAP70,IL2RG,JAK3,PNP,RFX5,PTPRC,IL7R,DCLRE1C(@>(CREBBP(3.99e@02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(CTDP1(3.26e@03);IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC(@>(REL(3.66e@02);ZAP70,IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(KDM6A(2.09e@02);ZAP70,IL2RG,RFX5,DCLRE1C(@>(ZBED4(2.61e@05);ADA,AK2,DCLRE1C(@>(CDK6(2.85e@04);ZA�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Specified(Hamartoses LUSC 0.63 PTEN 8.39E@06 PTEN(@>(skeletal(muscle(hypertrophy(is(regulated(via(akt@mtor(pathway(PIK3CA,PTEN);PTEN(@>(regulation(of(eif@4e(and(p70s6(kinase(PIK3CA,PTEN);PTEN(@>(mtor(signaling(pathway(PTEN,PIK3CA);PTEN(@>(Direct(p53(effectors(TP53,BID,EGFR,RB1,PTEN,CREBBP);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);PTEN(@>(RhoA(signaling(pathway(PTEN,MAPK12,SLC9A3);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN,PIK3CA);PTEN(@>(Signaling(events(mediated(by(Stem(cell(factor(receptor((c@Kit)(PIK3CA,CREBBP,PTEN)0.619582187 0 PTEN(@>(TTF2;SDHB(@>(EGFR;SDHD(@>(CDKN2A;STK11(@>(TP531Li(Fraumeni(and(Related(Syndromes LUSC 0.03 CDKN2A,TP53 5.79E@15 TP53(@>(Fluoropyrimidine(Pathway,(Pharmacodynamics(TYMS,TP53);TP53(@>(chaperones(modulate(interferon(signaling(pathway(TP53,RB1);TP53(@>(apoptotic(signaling(in(response(to(dna(damage(BID,TP53);TP53(@>(Direct(p53(effectors(TP53,BID,EGFR,RB1,PTEN,CREBBP);TP53,CDKN2A,CHEK2(@>(p53(pathway(CDKN2A,TP53,CREBBP);CDKN2A(@>(C@MYC(pathway(CDKN2A,FBXW7);TP53(@>(estrogen(responsive(protein(efp(controls(cell(cycle(and(breast(tumors(growth(TP53,CDK6);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(TP53,RB1);TP53(@>(btg(family(proteins(and(cell(cycle(regulation(TP53,RB1);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53(@>(p53(signaling(pathway(TP53,RB1);TP53(@>(regulation(of(transcriptional(activity(by(pml(TP53,CREBBP,RB1);TP53(@>(rb(tumor(suppressor/checkpoint(signaling(in(response(to(dna(damage(TP53,RB1);CDKN2A(@>(E2F(transcription(factor(network(TYMS,CDKN2A,CREBBP,RB1);TP53(@>(Glucocorticoid(receptor�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.651962919 0.02405 CDKN2A(@>(PTEN;CHEK2(@>(CDK6,TP53,PTEN;TP53(@>(CREBBP1Lipoprotein(Deficiencies LUSC 1 0.885354678 0.097427692 APOB,LCAT,APOA1(@>(ENOSF1(1.81e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(PROS1(1.81e@02);(@>(AKR1C2(2.16e@02);APOB,LCAT,APOA1(@>(SLC6A12(2.16e@02);MTTP,APOB,LCAT,APOA1(@>(SELO(1.87e@02)1 1Retinitis(Pigmentosa LUSC 1 EYS 1 5.69E@14 KLHL7,SPATA7,CRB1,MERTK,PDE6B,CERKL,PRCD,FAM161A(@>(CLCN4(4.22e@02);TTC8,CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,CRB1,NRL,PDE6B,PDE6G,CERKL,GUCA1B,TULP1,NR2E3,ROM1,USH2A,RP1,PRCD,IMPG2,C2orf71,SAG,RBP3,FAM161A(@>(EYS(3.56e@16);TTC8(@>(COLEC12(1.19e@03);TTC8,CRX,FSCN2,PRPH2,CNGB1,KLHL7,SPATA7,CRB1,MERTK,PDE6B,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(KCNIP4(1.35e@03);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(CLUL1(2.31e@14);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(SLC6A13(1.01e@12);TTC8,CRX,LRAT,FSCN2,RDH12,PRPH2,CNGB1,SPATA7,CRB1,MERTK,CERKL,TULP1,ROM1,PROM1,CNGA1,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(UNC13B(4.65e@06);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,RP1,PRCD,RPE65,SAG,RBP3,ABCA4(@>(KCNJ13(4.97e@12)1 1Haemophilia LUSC 1 0.00999936 F8,F9(@>(intrinsic(prothrombin(activation(pathway(COL4A5,PROS1)1 1 1Chronic(Granulomatous(Disease PRAD 1 1 0.078898011 NCF2,NCF4,CYBA(@>(GPS2(1.49e@02);NCF2,NCF4,CYBB,CYBA(@>(SPOPL(1.38e@02);NCF2,NCF4,CYBB,CYBA(@>(COTL1(3.89e@02);(@>(CRISPLD2(4.13e@02);NCF2,NCF4,CYBB,CYBA(@>(TMUB2(1.62e@02);NCF2(@>(MFI2(2.93e@02);CYBA(@>(ARRDC1(2.54e@02);NCF2,NCF4,CYBB,CYBA(@>(HELZ2(1.23e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.32e@02);NCF2,NCF4,CYBB(@>(PAK2(1.35e@02);NCF2,NCF4,CYBB,CYBA(@>(NFATC1(1.34e@02);NCF2,NCF4,CYBB,CYBA(@>(CRTC2(1.42e@02);NCF2,NCF4,CYBB,CYBA(@>(ADRBK1(1.57e@02);NCF2,NCF4,CYBA(@>(ZGPAT(1.40e@02);NCF2,NCF4(@>(EGR3(2.34e@02);NCF2,NCF4,CYBA(@>(GPR160(1.26e@02);NCF2,CYBB,CYBA(@>(HNMT(1.21e@02);NCF2,NCF4,CYBB(@>(SENP5(3.88e@02);NCF2,NCF4,CYBB,CYBA(@>(DNAJC5(1.35e@02);NCF2,NCF4,CYBB,CYBA(@>(ANKRD13D(1.27e@02);NCF2,NCF4,CYBB,CYBA(@>(CTDP1(1.30e@02);NCF4,CYBA(@>(CDKN1B(3.93e@02);CYBA(@>(POLD4(2.93e@02);CYBA(@>(RPS27(2.87e@02);NCF2,NCF4,CYBB,CYBA(@>(TPD52L2(1.12e@02);NCF2,NCF4,CYBA(@>(GMEB2(1.40e@02);NCF2,NCF4,CYBB,CYBA(@>(PQLC1(1.16e@02);NCF2,NCF4,CYBB,CYBA(@>(RGS19(2.46e@02);NCF2,NCF4,CYBB,CYBA(@>(TOR4A(1.38e@02);NCF2,NCF4,CYBB(@>(R�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Glycogenosis PRAD 1 1 0.078728665 PGAM2,PHKB,PHKA2,AGL(@>(SLC25A30(1.28e@02)1 1Congenital(Ichthyosis PRAD 1 1 0.02654965 ALOX12B,SPINK5,CSTA,TGM1(@>(C9orf169(2.49e@03);ALOX12B,ALOXE3,CSTA,TGM1(@>(PARD6G(3.82e@02);ALOXE3,CSTA,ABCA12,TGM1(@>(ARRDC1(2.44e@02);NIPAL4,KRT2,LIPN,ABCA12(@>(NRARP(4.60e@03);SPINK5,CSTA,NIPAL4,ABCA12,TGM1(@>(PTK6(4.60e@03)1 1Disorders(of(Phosphorous(Metabolism PRAD 0.69 SLC34A3 0.000406927 SLC34A3(@>(Type(II(Na+/Pi(cotransporters(SLC34A3) 1 1 1Spinocerebellar(Ataxia PRAD 1 1 0.027197561 APTX,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,SYT14,C10orf2,PPP2R2B(@>(CASP8AP2(3.80e@03);JPH3,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1(@>(CREBL2(1.97e@02);JPH3,APTX,CACNA1A,ATXN2,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@>(DLG1(5.25e@03);JPH3,TDP1,KCNC3,FGF14(@>(SAMD10(1.32e@02);JPH3,CACNA1A,ATXN2,PDYN,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,SYT14,PPP2R2B(@>(NSMF(9.90e@03);JPH3,ZNF592,CACNA1A,ATXN7,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,FGF14,SETX,PPP2R2B,ATXN1(@>(PHC3(2.61e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(GRIN1(3.68e@03);JPH3,CACNA1A,ATXN2,SYNE1,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(MYT1(2.84e@03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B(@>(ZNF285(1.11e@02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(STMN3(2.84e@03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Specified(Hamartoses PRAD 0.66 PTEN 0.062379972 STK11(@>(Metformin(Pathway,(Pharmacodynamic(SLC2A4,CRTC2);PTEN(@>(RhoA(signaling(pathway(CDKN1B,PTEN);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(CDKN1B,PTEN);PTEN(@>(Negative(regulation(of(the(PI3K/AKT(network(PTEN)0.57662662 1 1Retinitis(Pigmentosa PRAD 1 1 6.10E@15 CRX,EYS,SPATA7,CRB1,MERTK,PDE6B,CERKL,PRCD,IMPG2,C2orf71,RBP3(@>(CREBL2(4.56e@02);CRX,FSCN2,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,PDE6B,CERKL,TULP1,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(MYT1(7.66e@06);SNRNP200,CRB1,PRPF31,CERKL,IDH3B,PRPF8,FAM161A(@>(UCKL1(1.68e@03);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,EYS,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(KCNG4(2.10e@17);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,EYS,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4(@>(SAMD7(1.45e@17);RPGR,CA4,PRPF3,BEST1,RP2,TOPORS,SEMA4A(@>(GPR160(1.80e@02);SNRNP200,EYS(@>(THSD7B(1.89e@05);(@>(WFDC1(1.49e@05);(@>(NXPH2(5.64e@03);TTC8,RDH12,SPATA7,MERTK,CNGA1,FAM161A(@>(ZFHX3(1.68e@03);CRX,FSCN2,RDH12,SNRNP200,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,PDE6B,CERKL,PRPF3,TULP1,ROM1,PRCD,IMPG2,C2orf71,TOPORS,RBP3,FAM161A(@>(PCMTD2(7.66e@06);ZNF513,KLHL7,SPATA7,CRB1,PDE6B,CERKL,FAM161A(@>(TMEM145(2.�������0.34981818 0.26722222Long(QT(Syndrome READ 0.69 CACNA1C 3.14E@21 CACNA1C(@>(Nicotine(Pathway((Chromaffin(Cell),(Pharmacodynamics(CACNA1C);CACNA1C(@>(Sympathetic(Nerve(Pathway((Pre@(and(Post@(Ganglionic(Junction)(CACNA1C,TH);CACNA1C(@>(Anti@diabetic(Drug(Potassium(Channel(Inhibitors(Pathway,(Pharmacodynamics(PDX1,CACNA1C,INS)1 1 1Chronic(Granulomatous(Disease READ 1 1 0.07714767 (@>(CRLF2(3.23e@02);NCF2,NCF4,CYBB,CYBA(@>(PRPF3(2.17e@02);NCF2,NCF4,CYBB,CYBA(@>(CACNA2D4(1.29e@02);NCF2(@>(KRAS(4.79e@02);NCF4,CYBA(@>(TRAPPC2(2.54e@02);NCF2,NCF4,CYBB,CYBA(@>(PLAGL2(1.33e@02);NCF2,NCF4,CYBB,CYBA(@>(CSF2RA(1.45e@02);NCF2,CYBB,CYBA(@>(TCF7L2(1.37e@02);NCF2,NCF4,CYBA(@>(TCEANC(2.36e@02);NCF2,NCF4,CYBB(@>(CTNNBL1(1.42e@02);NCF2,NCF4(@>(RNF40(3.72e@02);NCF2,NCF4,CYBB,CYBA(@>(IRF2(1.53e@02);NCF2,NCF4,CYBB,CYBA(@>(CR1(1.22e@02);NCF2,NCF4,CYBB,CYBA(@>(FBRS(1.23e@02);NCF2,CYBB,CYBA(@>(HS3ST3B1(1.16e@02);NCF4(@>(C17orf103(2.44e@02);NCF2,NCF4,CYBB,CYBA(@>(IL3RA(1.69e@02);(@>(ADK(3.11e@02);NCF2,CYBB(@>(RAB9A(1.57e@02);NCF2,CYBB(@>(EMP1(4.04e@02);NCF2(@>(SRCAP(3.53e@02);NCF2,NCF4,CYBB,CYBA(@>(RAB39A(1.60e@02);NCF2,NCF4,CYBB,CYBA(@>(MAP2K3(2.17e@02);NCF2,NCF4,CYBB,CYBA(@>(MOSPD2(1.16e@02);NCF2,NCF4,CYBB,CYBA(@>(MCL1(1.19e@02);NCF2(@>(P2RY8(1.27e@02)1 1Glycogenosis READ 0.86 PHKG2 6.77E@12 AGL,PYGM,PHKA1,PHKB,PHKA2,PHKG2(@>(Glycogen(breakdown((glycogenolysis)(PHKG2,GYG2)0.372442654 1 0.36782353Inherited(Anomalies(of(the(Skin READ 1 0.004631074 TERT(@>(HIF@1@alpha(transcription(factor(network(SMAD4,MCL1,SMAD3);TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(TP53,KRAS);TERT(@>(Validated(targets(of(C@MYC(transcriptional(activation(SMAD4,TP53,SMAD3)0.10458489 1 1Spinocerebellar(Ataxia READ 1 ATXN10 0.008075087 CACNA1A(@>(Sympathetic(Nerve(Pathway((Pre@(and(Post@(Ganglionic(Junction)(CACNA1C,TH);PRNP(@>(Glypican(1(network(HCK,FGFR1);CACNA1A(@>(Anti@diabetic(Drug(Potassium(Channel(Inhibitors(Pathway,(Pharmacodynamics(PDX1,CACNA1C,INS);TBP(@>(Validated(targets(of(C@MYC(transcriptional(repression(ERBB2,SMAD4,SMAD3)0.013136049 ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,PPP2R2B(@>(FHIT(3.07e@03);(@>(PPP2R3B(4.28e@02);JPH3,ATXN2,TTBK2,TBP,PPP2R2B(@>(CDRT4(4.76e@02);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B(@>(GPM6B(3.44e@03);ZNF592,POLG,ATXN7,TDP1,ATM,TBP,ITPR1,SETX,ATXN1(@>(PRPF3(8.48e@03);JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1(@>(CCSER1(2.41e@03);JPH3,ZNF592,CACNA1A,ATXN7,ATXN2,PDYN,TDP1,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B,ATXN1(@>(WHSC1L1(1.11e@03);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2B(@>(PARK2(2.44e@03);JPH3,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,PPP2R2B(@>(ENSA(4.21e@02);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1(@>(ZNF785(1.51e@03);JPH3,ZNF592,CACNA1A,ATXN7,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,FGF14,SETX,SYT14,PPP2R2B,ATXN1(@>(CUL5(1.42e@03);POLG,ATXN7,TDP1,SYNE1,ATM,TBP,ITPR1,�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Severe(Combined(Immunodeficiency READ 1 1 0.017157155 IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,DCLRE1C(@>(PRPF3(1.29e@02);ZAP70,IL2RG,CIITA,JAK3,RFXAP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(OFD1(1.32e@03);RFXAP,DCLRE1C(@>(FNTA(4.56e@02);PNP(@>(CACNA2D4(4.19e@02);ZAP70,IL2RG,JAK3,RFXAP,IL7R,CD3D,DCLRE1C(@>(CSTF2T(1.10e@02);(@>(DDX47(3.75e@02);IL2RG,JAK3,RFXANK,RFX5,PTPRC(@>(PRR14(6.34e@03);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,DCLRE1C(@>(PLAGL2(1.07e@02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(TCEANC(2.91e@03);ZAP70,IL2RG,JAK3,RFXAP,RFX5,IL7R,CD3D,DCLRE1C(@>(ASXL1(1.37e@03);PNP(@>(CTNNBL1(3.18e@02);IL2RG,JAK3,ADA,PNP,PTPRC(@>(HCK(4.98e@02);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,DCLRE1C(@>(IRF2(9.30e@03);IL2RG,JAK3,ADA,PNP,PTPRC(@>(HS3ST3B1(1.89e@02);JAK3,PTPRC(@>(C17orf103(4.06e@02);RFX5(@>(GPRC5D(1.63e@02);CIITA(@>(FLT3(1.41e@03);RFXAP,AK2,DCLRE1C(@>(FANCB(2.62e@02);CIITA,RFX5,DCLRE1C(@>(ADK(1.46e@02);PNP,PTPRC(@>(P2RY8(2.19e@03);RFXAP(@>(ANK1(4.68e@02)1 1Lipoprotein(Deficiencies READ 1 1 0.075339054 MTTP,APOB,LCAT,SAR1B,APOA1(@>(KLC4(3.60e@02);APOB,LCAT,APOA1(@>(FCN3(2.07e@02);MTTP(@>(INS@IGF2(2.31e@02);APOB,ABCA1(@>(HS3ST3B1(2.70e@02);(@>(ADK(3.79e@02);MTTP,APOB,LCAT,APOA1(@>(ARSD(1.18e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ARSE(2.07e@02)1 1Disorders(of(Urea(Cycle(Metabolism READ 1 1 0.088390959 ASS1,NAGS,ARG1,ASL,CPS1(@>(KLC4(4.33e@02);NAGS,ARG1,ASL,CPS1(@>(FCN3(2.80e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ARSD(1.57e@02);ASS1,NAGS,ARG1,ASL,CPS1(@>(ARSE(2.80e@02)1 1Retinitis(Pigmentosa READ 1 PRPF3 1 4.04E@13 (@>(IMMP2L(2.15e@05);(@>(DHRSX(1.34e@03);EYS,CERKL(@>(NKX6@3(1.42e@02);RPGR,CRX,FSCN2,RDH12,PRPH2,CNGB1,EYS,CRB1,TULP1,ROM1,RP2,PRCD,IMPG2,C2orf71,RBP3,SEMA4A(@>(CACNA2D4(1.12e@08);FSCN2,PRPH2,RHO,CNGB1,PDE6A,PDE6G,GUCA1B,RLBP1,RGR,NR2E3,ROM1,CNGA1,RP1,SAG,ABCA4(@>(EGFL6(4.38e@09);CRX,FSCN2,RDH12,PRPH2,CNGB1,EYS,CRB1,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3(@>(ASMT(1.12e@08);TTC8,CRX,FSCN2,RDH12,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,MERTK,PDE6B,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(KIAA1467(6.02e@04);(@>(ZBED1(3.70e@03);TTC8,CRX,FSCN2,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,TULP1,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(GSG1(6.99e@07);CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6G,GUCA1B,TULP1,NR2E3,ROM1,PROM1,CNGA1,RP1,PRCD,C2orf71,SAG,RBP3,FAM161A,ABCA4(@>(KERA(5.05e@15);FAM161A(@>(SHOX(9.94e@11)1 0.40083333Dopa@Responsive(Dystonia READ 0.62 TH 2.34E@32 TH(@>(Alpha@synuclein(signaling(HCK,TH,PARK2);TH(@>(Sympathetic(Nerve(Pathway((Pre@(and(Post@(Ganglionic(Junction)(CACNA1C,TH)1 1 1Congenital(Ectodermal(Dysplasia SKCM 1 0.28548436 0.000106457 (@>(FAM58A(9.62e@03);LAMB3,ITGB4,PLEC,KRT5,LAMC2(@>(ANAPC15(2.97e@03);(@>(EIF3D(1.58e@03);ITGA6,LAMB3,ITGB4,GJB6,COL17A1,PLEC,COL7A1,KRT5,KRT14,LAMC2,LAMA3(@>(PTK6(2.00e@06);LAMB3,ITGB4,KRT5(@>(MYC(2.50e@04);(@>(CDKN2A(2.84e@02);PLEC(@>(PHGDH(3.27e@02);PLEC(@>(RPL13(4.57e@03);(@>(CCND1(2.48e@02);LAMB3,ITGB4,PLEC,KRT5(@>(PPDPF(2.47e@03);(@>(LSM12(2.57e@02);(@>(SLC25A39(1.42e@02);(@>(MRP63(4.21e@02);(@>(SRMS(2.88e@02);ITGA6,LAMB3,ITGB4,COL17A1,KRT5,KRT14,LAMC2,LAMA3(@>(TDRP(5.01e@04);PLEC(@>(GRN(1.79e@02);(@>(TP53(1.56e@02);(@>(TSPAN31(1.19e@02);(@>(HDAC3(3.26e@03);ITGB4,PLEC,KRT5(@>(ACD(5.04e@03);(@>(KRT78(1.13e@02);GJB6(@>(TCHHL1(9.83e@03);PLEC(@>(DEF8(2.72e@02);ITGA6,COL17A1,COL7A1,KRT14(@>(DSG1(1.08e@05);LAMB3,ITGB4,PLEC,KRT5,KRT14,LAMC2(@>(TNFRSF6B(6.12e@04);(@>(RPTN(3.38e@04);PLEC(@>(TPD52L2(4.98e@02);LAMB3,ITGB4,PLEC,COL7A1,KRT5(@>(CHMP1A(1.44e@02);(@>(DYNAP(9.82e@03);ITGB4,PLEC,COL7A1(@>(SLC2A4RG(6.70e@03);(@>(FOLR3(2.43e@03)1 1Chronic(Granulomatous(Disease SKCM 1 1 0.078763609 NCF2,NCF4,CYBB,CYBA(@>(ZFX(1.54e@02);NCF2,NCF4,CYBB,CYBA(@>(HELZ2(1.79e@02);NCF2,NCF4(@>(KIAA1257(2.33e@02);NCF2(@>(STK19(3.34e@02);NCF2,NCF4,CYBB,CYBA(@>(B2M(2.08e@02);NCF2,NCF4,CYBB(@>(VPS9D1(1.84e@02);NCF2,NCF4,CYBB,CYBA(@>(DDX3X(1.31e@02);NCF2,NCF4,CYBB,CYBA(@>(TMUB2(1.76e@02);NCF2,NCF4,CYBB,CYBA(@>(CLEC2B(2.70e@02);NCF2,NCF4,CYBB,CYBA(@>(ADAM8(1.31e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.49e@02);NCF2,NCF4,CYBB,CYBA(@>(ZNF276(1.43e@02);NCF4(@>(SLC25A39(1.56e@02);NCF2,NCF4,CYBB,CYBA(@>(GNAI2(1.51e@02);NCF2,NCF4,CYBB,CYBA(@>(PPP6C(1.49e@02);NCF2,NCF4,CYBB,CYBA(@>(RPGRIP1(1.45e@02);NCF4,CYBB(@>(POM121(3.52e@02);NCF2,NCF4,CYBA(@>(ZGPAT(1.56e@02);(@>(SERPINB10(1.35e@02);NCF2,NCF4,CYBB,CYBA(@>(GRN(3.32e@02);NCF2,NCF4,CYBB,CYBA(@>(DNAJC5(1.52e@02);NCF2,NCF4,CYBB,CYBA(@>(RBM22(1.44e@02);NCF4(@>(ITGA2B(3.33e@02);CYBB,CYBA(@>(OXA1L(1.33e@02);NCF2,NCF4,CYBB(@>(MPP7(1.54e@02);(@>(SLC4A1(3.39e@02);NCF2,NCF4,CYBB,CYBA(@>(CLEC5A(1.39e@02);NCF2,NCF4,CYBB,CYBA(@>(GPR141(1.49e@02);NCF4,CYBB,CYBA(@>(ITGA4(1.49e@02);NCF2,NCF4,CYB������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Congenital(Ichthyosis SKCM 1 1 0.005136028 SPINK5,CSTA,NIPAL4,ABCA12,TGM1(@>(PTK6(3.02e@03);SPINK5,CSTA,NIPAL4,KRT2,LIPN,ABHD5(@>(MPP7(1.32e@02);ALOX12B,SPINK5,CSTA,TGM1(@>(KRT78(5.11e@03);ALOX12B,ABCA12(@>(TCHHL1(3.36e@04);ALOX12B,SPINK5,CSTA,ABCA12,TGM1(@>(DSG1(2.89e@04);ALOX12B,SPINK5,CSTA,ABCA12,TGM1(@>(RPTN(2.89e@04);ALOX12B,SPINK5,CSTA,TGM1(@>(DYNAP(3.02e@03)1 1Polycystic(Kidney,(Autosomal(Dominant SKCM 1 0.013304666 TSC2(@>(LKB1(signaling(events(RPTOR,MYC,TP53);TSC2(@>(Validated(targets(of(C@MYC(transcriptional(repression(CCND1,HDAC3,EP300,MYC)0.736679678 1 1Inherited(Anomalies(of(the(Skin SKCM 1 TERT 5.99E@19 TERT(@>(erk1/erk2(mapk(signaling(pathway(MYC,TERT);TERT(@>(IL2(signaling(events(mediated(by(PI3K(MYC,TERT,RAC1);TERT(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(MYC,TERT,TP53);TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(MYC,TERT,TP53);KRT1,TERT(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);TERT(@>(role(of(nicotinic(acetylcholine(receptors(in(the(regulation(of(apoptosis(TERT,FASLG);TERT(@>(Validated(targets(of(C@MYC(transcriptional(activation(TERT,TP53,UBTF,CDK4,MYC,EP300);TINF2,TERT,DKC1(@>(Regulation(of(Telomerase(CCND1,MYC,TERT,SAP18,ACD)0.164716031 1 1Combined(Heart(and(Skeletal(Defects SKCM 0.64 EP300 2.01E@19 EP300(@>(cell(cycle:(g2/m(checkpoint(EP300,TP53,MYT1);EP300(@>(Validated(transcriptional(targets(of(AP1(family(members(Fra1(and(Fra2(CCND1,EP300,CDKN2A);EP300,CREBBP(@>(acetylation(and(deacetylation(of(rela(in(nucleus(HDAC3,EP300);EP300(@>(Notch(signaling(pathway(CCND1,MYC,EP300,NOTCH2,DNER);EP300(@>(Validated(nuclear(estrogen(receptor(alpha(network(CCND1,KLRC3,EP300,MYC);EP300(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);EP300(@>(Validated(targets(of(C@MYC(transcriptional(repression(CCND1,HDAC3,EP300,MYC);EP300,CREBBP(@>(E2F(transcription(factor(network(CCNE2,EP300,MYC,CDKN2A);EP300,CREBBP(@>(il@7(signal(transduction(ITGA2B,EP300);EP300,CREBBP(@>(Validated(targets(of(C@MYC(transcriptional(activation(TERT,TP53,UBTF,CDK4,MYC,EP300);EP300(@>(hypoxia(and(p53(in(the(cardiovascular(system(EP300,TP53);EP300(@>(ATF@2(transcription(factor(network(CCND1,EP300,CDK4,NF1);EP300(@>(melanocyte(development(and(pigmentation(pathway(MITF,EP300);EP300,CREBBP(@>����������������������������������������������������������������1 0.32066667 1Specified(Anomalies(of(the(Musculoskeletal(SystemSKCM 1 MITF 0.097075171 MITF,SNAI2(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);MITF(@>(IL6@mediated(signaling(events(MITF,MYC,RAC1);MITF(@>(melanocyte(development(and(pigmentation(pathway(MITF,EP300)1 1 1Neurofibromatosis SKCM 0.61 NF1 4.53E@05 NF1(@>(Regulation(of(Ras(family(activation(NRAS,RASA2,NF1);NF1(@>(ATF@2(transcription(factor(network(CCND1,EP300,CDK4,NF1)0.719411434 0.222 NF2(@>(CDKN2A 1Tuberous(Sclerosis SKCM 1 0.013304666 TSC2,TSC1(@>(LKB1(signaling(events(RPTOR,MYC,TP53);TSC2(@>(Validated(targets(of(C@MYC(transcriptional(repression(CCND1,HDAC3,EP300,MYC)0.594661126 1 1Severe(Combined(Immunodeficiency SKCM 1 1 0.001386824 IL2RG,JAK3,ADA,PNP,PTPRC(@>(HELZ2(4.66e@02);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1(@>(THEMIS(1.76e@04);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(B2M(6.45e@03);NHEJ1,RFX5,AK2(@>(EIF3D(4.81e@05);ZAP70,IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(KCNAB2(3.23e@03);ZAP70,RFXANK,IL7R,CD3D(@>(RPL13(2.74e@03);AK2(@>(TERT(1.68e@02);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C(@>(TC2N(5.55e@03);ZAP70,IL2RG,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(CLEC2B(3.84e@03);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(CLEC2D(1.81e@04);IL2RG,JAK3,PNP,RFX5,PTPRC(@>(ADAM8(3.98e@02);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(ZNF276(3.22e@03);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(PCED1B(1.76e@03);JAK3(@>(SLC25A39(7.97e@04);IL2RG,CIITA,PNP,RFX5,PTPRC,DCLRE1C(@>(LY86(1.75e@02);ZAP70,IL2RG,JAK3,PNP,RFX5,PTPRC,DCLRE1C(@>(PPP6C(5.65e@03);ZAP70,IL2RG,JAK3,RFXAP,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C(@>(POM121(8.56e@04);ZAP70,IL2RG,JAK3,ADA,PTPRC,IL7R,CD3D(@>(ZGPAT(2.74e@03);PNP,AK2(@>(����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Specified(Hamartoses SKCM 0.72 PTEN 0.061006438 STK11(@>(LKB1(signaling(events(RPTOR,MYC,TP53);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);STK11(@>(Regulation(of(AMPK(activity(via(LKB1(RPTOR);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN,FASLG)0.561087039 0.13528125 PTEN(@>(TP53;SDHB(@>(CDKN2A;STK11(@>(NRAS,OGDHL,TP530.2405 PTEN(@>(TCEB3C;SDHB(@>(TP53;STK11(@>(CCNE2;VHL(@>(CHGBLi(Fraumeni(and(Related(Syndromes SKCM 0.05 CDKN2A,TP53 2.63E@31 TP53(@>(LKB1(signaling(events(RPTOR,MYC,TP53);TP53,CHEK2(@>(cell(cycle:(g2/m(checkpoint(EP300,TP53,MYT1);CDKN2A(@>(Validated(transcriptional(targets(of(AP1(family(members(Fra1(and(Fra2(CCND1,EP300,CDKN2A);TP53(@>(estrogen(responsive(protein(efp(controls(cell(cycle(and(breast(tumors(growth(CDK4,TP53);TP53(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(MYC,TERT,TP53);TP53,CHEK2(@>(role(of(brca1(brca2(and(atr(in(cancer(susceptibility(TP53,FANCA);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(MYC,TERT,TP53);CDKN2A(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);TP53(@>(btg(family(proteins(and(cell(cycle(regulation(CCND1,TP53);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53(@>(p53(signaling(pathway(CCND1,CDK4,TP53);TP53(@>(rb(tumor(suppressor/checkpoint(signaling(in(response(to(dna(damage(CDK4����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.662892321 0.15561765 CDKN2A(@>(PTEN;CHEK2(@>(CDK4,TP53,PTEN;TP53(@>(EP3001Lipoprotein(Deficiencies SKCM 1 1 0.058815075 APOB,LCAT,APOA1(@>(SPTLC3(3.68e@02);APOB,SAR1B(@>(IDH1(2.34e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(UGT2B15(6.62e@03);MTTP,APOB,LCAT,SAR1B,APOA1,ABCA1(@>(C2(1.13e@02);APOB,APOA1(@>(TIMD4(3.68e@02);APOB,APOA1(@>(STAB2(1.22e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(APCS(1.13e@02)1 1Retinitis(Pigmentosa SKCM 1 EYS,CRB1 1 2.00E@13 CERKL,IDH3B(@>(ASB16(5.43e@03);CRX,FSCN2,KLHL7,SPATA7,MERTK,PDE6B,CERKL,ROM1,PRCD,C2orf71,RBP3,FAM161A(@>(SV2B(3.21e@03);CRX,FSCN2,PRPH2,CNGB1,KLHL7,SPATA7,PDE6B,CERKL,TULP1,PRCD,IMPG2,C2orf71,RBP3,FAM161A(@>(MYT1(4.84e@05);TTC8,CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,KLHL7,SPATA7,NRL,MERTK,PDE6B,PDE6G,CERKL,GUCA1B,TULP1,NR2E3,ROM1,PRCD,IMPG2,C2orf71,SAG,RBP3,FAM161A,ABCA4(@>(CRB1(2.61e@05);CRX,FSCN2,RDH12,PRPH2,CNGB1,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3(@>(LRTM1(5.49e@09);TTC8,CRX,FSCN2,RDH12,PRPH2,CNGB1,SPATA7,PDE6B,TULP1,ROM1,PRCD,RBP3(@>(FMN1(2.31e@04);CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,PROM1,CNGA1,RP1,PRCD,C2orf71,SAG,RBP3,FAM161A,ABCA4(@>(PROL1(6.23e@14);KLHL7(@>(NUDT4(2.32e@02);KLHL7,IDH3B(@>(SCN5A(4.86e@02);LRAT,SNRNP200,PROM1(@>(UTF1(4.41e@02);ZNF513,KLHL7,SPATA7,MERTK,PDE6B,PRCD,FAM161A(@>(CHRNA4(1.59e@02);CRX,SNRNP200,CA4,CERKL,TULP1,PROM1,FAM161A(@>(KCNB2(9.21e@04);TTC8,RPGR,SPATA7,CERKL,FAM161A(@>(CHGB(4.86e@02);(@>(MITF(1.67e@04);(@>(SPTBN5(4.63e@�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.21645 IDH3B(@>(IDH1,EIF2B1;IMPDH1(@>(SKIV2L;PROM1(@>(DDX3X;PRPF3(@>(RPL13,CRB1,PRPF6,PRPF6,DDX3X;PRPF31(@>(PRPF6,RPGRIP1,MYC;PRPF8(@>(PRPF60.22446667 CA4(@>(SLC4A1,EP300;CRX(@>(RBFOX1;FAM161A(@>(PARK2,RUNDC3A,MYC;IDH3B(@>(TP53,DDX3X,HDAC3,PRPF6,FANCA;IMPDH1(@>(PRPF6,DDX3X;NR2E3(@>(ITGA4;PRPF3(@>(HDAC5;PRPF31(@>(MYC;PRPF8(@>(PRPF6,PHGDH,RPGRIP1,HDAC5;ROM1(@>(ITGA4;RPGR(@>(PARK2;SNRNP200(@>(MYC;TOPORS(@>(PRPF6,UPF3AHereditary(Hemorrhagic(Telangiectasia SKCM 1 1.86E@06 SMAD4(@>(LKB1(signaling(events(RPTOR,MYC,TP53);SMAD4(@>(Validated(nuclear(estrogen(receptor(alpha(network(CCND1,KLRC3,EP300,MYC);SMAD4(@>(Validated(targets(of(C@MYC(transcriptional(repression(CCND1,HDAC3,EP300,MYC);SMAD4(@>(Validated(targets(of(C@MYC(transcriptional(activation(TERT,TP53,UBTF,CDK4,MYC,EP300)1 1 1Disorders(of(Aromatic(Amino(Acid(MetabolismSKCM 1 MC1R 1 0.072585714 BLOC1S6,AP3B1,BLOC1S3,HPS5,HPS1(@>(HELZ2(1.61e@02);BLOC1S6,AP3B1,BLOC1S3,HPS5,HPS6,HPS1(@>(TMUB2(3.76e@02);TYR,OCA2,TYRP1,SLC45A2,BLOC1S6,AP3B1,DTNBP1,BLOC1S3,HPS5,HPS1(@>(KCNAB2(9.33e@03);BLOC1S6,AP3B1,BLOC1S3,HPS6,HPS1(@>(GNAI2(2.33e@02);BLOC1S6,BLOC1S3,HPS5,HPS6,HPS1(@>(GRN(1.01e@02);(@>(OXA1L(2.24e@02);BLOC1S6,AP3B1,HPS5(@>(CLEC5A(3.63e@02);BLOC1S6,BLOC1S3,HPS5,HPS6,HPS1(@>(TPD52L2(3.55e@02);TAT,HPD,FAH(@>(C2(3.91e@02);HPD(@>(LIME1(4.55e@02);BLOC1S6,AP3B1,BLOC1S3,HPS5,HPS1(@>(RGS19(3.70e@02);TYR,OCA2,TYRP1,SLC45A2(@>(MITF(2.04e@02);BLOC1S3,HPS6,HPS1(@>(INPPL1(2.36e@02);TYR,OCA2,TYRP1,SLC45A2,BLOC1S3,HPS5,HPS1(@>(MAD1L1(1.01e@02);BLOC1S6,DTNBP1,BLOC1S3,HPS1(@>(TCF25(3.81e@02);BLOC1S6,AP3B1,DTNBP1,BLOC1S3,HPS5,HPS1(@>(SBNO2(2.33e@02);HPS1(@>(FOLR2(1.98e@02)1 1Chronic(Granulomatous(Disease STAD 1 0.846838069 0.072585714 NCF2,NCF4,CYBB,CYBA(@>(B2M(2.04e@02);NCF2,NCF4,CYBB,CYBA(@>(DIAPH2(1.14e@02);NCF2,NCF4,CYBB,CYBA(@>(TRPS1(9.36e@03);NCF2(@>(KRAS(4.90e@02);CYBA(@>(CLECL1(3.26e@02);NCF2,NCF4,CYBB,CYBA(@>(CLEC2B(2.77e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.14e@02);NCF2,NCF4,CYBB,CYBA(@>(SNX2(1.75e@02);NCF2,NCF4,CYBB,CYBA(@>(CLEC12A(1.60e@02);NCF4(@>(IL5RA(3.13e@02);NCF2,CYBB,CYBA(@>(RHOA(9.22e@03);NCF2,NCF4,CYBB,CYBA(@>(IRF2(9.69e@03);(@>(KLRF1(3.81e@02);(@>(CLEC1B(3.30e@02);NCF4,CYBA(@>(CD69(3.41e@02);NCF2,NCF4,CYBB,CYBA(@>(DYRK1A(2.15e@02);NCF2,CYBB,CYBA(@>(UAP1L1(1.35e@02);(@>(PLGRKT(3.41e@02);CYBB,CYBA(@>(DPP7(1.97e@02);NCF2,CYBB,CYBA(@>(CD44(9.47e@03);NCF2,NCF4,CYBB,CYBA(@>(CD274(1.06e@02)1 1Disorders(of(Phosphorous(Metabolism STAD 0.67 SLC34A3 0.003812225 FGF23(@>(Syndecan@2@mediated(signaling(events(RHOA,FGFR2,FGF19);FGF23(@>(Syndecan@3@mediated(signaling(events(FGFR2,FGF19,EGFR);SLC34A3(@>(Type(II(Na+/Pi(cotransporters(SLC34A3);FGF23(@>(FGF(signaling(pathway(CDH1,PIK3CA,FGF19,FGFR2)1 0.11223333 FGF23(@>(FGF19 1Combined(Heart(and(Skeletal(Defects STAD 1 0.002558855 CREBBP(@>(the(information(processing(pathway(at(the(ifn(beta(enhancer(IRF2,ARID1A);EP300(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDH1,APC,AES,CDKN2A);EP300(@>(p73(transcription(factor(network(RNF43,CDK6,WWOX)0.650459398 0.04182609 CREBBP(@>(TP53,TP53,GATA40.26054167Specified(Hamartoses STAD 0.63 PTEN 1.38E@07 PTEN(@>(skeletal(muscle(hypertrophy(is(regulated(via(akt@mtor(pathway(PIK3CA,PTEN);PTEN(@>(regulation(of(eif@4e(and(p70s6(kinase(PIK3CA,PTEN);PTEN(@>(mtor(signaling(pathway(PTEN,PIK3CA);VHL(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);PTEN(@>(RhoA(signaling(pathway(PTEN,RHOA,MAP2K4);PTEN(@>(pten(dependent(cell(cycle(arrest(and(apoptosis(PTEN,PIK3CA);PTEN(@>(Negative(regulation(of(the(PI3K/AKT(network(PTEN)0.719559223 0 PTEN(@>(PIK3CA;STK11(@>(EGFR1Li(Fraumeni(and(Related(Syndromes STAD 0.03 CDKN2A,TP53 1.68E@20 CDKN2A(@>(C@MYC(pathway(CDKN2A,FBXW7);TP53(@>(estrogen(responsive(protein(efp(controls(cell(cycle(and(breast(tumors(growth(TP53,CDK6);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(CDKN2A,TP53);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(TP53,KRAS);CDKN2A(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CDH1,APC,AES,CDKN2A);TP53(@>(BARD1(signaling(events(CCNE1,TP53);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53,CHEK2(@>(PLK3(signaling(events(CCNE1,TP53);TP53(@>(p53(signaling(pathway(CCNE1,TP53);TP53(@>(p75(NTR)@mediated(signaling(PIK3CA,RHOA,TP53)0.693511719 0 CDKN2A(@>(PTEN;CHEK2(@>(CDK6,TP53,SMAD4;TP53(@>(PTEN1Chronic(Granulomatous(Disease UCEC 1 1 0.078663977 NCF2,NCF4,CYBA(@>(GMEB2(1.50e@02);NCF2,NCF4,CYBA(@>(ZNF263(1.79e@02);NCF2,CYBA(@>(IRAK1(2.62e@02);NCF2,NCF4,CYBB,CYBA(@>(HELZ2(1.44e@02);NCF2(@>(KRAS(4.41e@02);NCF2,NCF4,CYBB,CYBA(@>(ADAMDEC1(4.66e@02);NCF2,NCF4,CYBB,CYBA(@>(PTEN(1.31e@02);NCF2,NCF4,CYBB,CYBA(@>(PLAGL2(1.18e@02);NCF2,NCF4,CYBB,CYBA(@>(NFATC1(1.48e@02);NCF2,NCF4,CYBB,CYBA(@>(VDR(1.25e@02);NCF2,NCF4,CYBB,CYBA(@>(DNM2(1.64e@02);(@>(HAUS8(4.54e@02);NCF2,NCF4,CYBA(@>(ZGPAT(1.41e@02);NCF4,CYBB,CYBA(@>(NEK8(1.32e@02);NCF2,NCF4,CYBB,CYBA(@>(NFE2L2(1.19e@02);NCF2,NCF4,CYBB,CYBA(@>(CREBBP(1.30e@02);CYBA(@>(TMEM80(2.60e@02);NCF2,NCF4,CYBB,CYBA(@>(CTDP1(1.26e@02);NCF2,NCF4,CYBA(@>(CHMP2A(1.31e@02);CYBA(@>(RPLP2(4.66e@02);NCF2,NCF4,CYBB,CYBA(@>(SAP30BP(1.29e@02);CYBA(@>(POLD4(2.83e@02);NCF2,NCF4,CYBB,CYBA(@>(TPD52L2(1.15e@02);NCF4,CYBB,CYBA(@>(ADAM28(1.19e@02);NCF2,NCF4,CYBB,CYBA(@>(PQLC1(1.18e@02);NCF2,CYBB(@>(CTNNB1(3.35e@02);NCF2,NCF4,CYBB,CYBA(@>(SGK1(1.18e@02);NCF2,NCF4,CYBB,CYBA(@>(MYO9B(1.42e@02);NCF2,CYBB,CYBA(@>(NEU4(1.31e@02);NCF2,NCF4,CYBB,CYBA���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1Diamond@Blackfan(Anemia UCEC 1 1.14E@06 RPS26,RPS24,RPS10,RPS17,RPS19,RPL5,RPL35A,RPS7,RPL11(@>(Regulation(of(gene(expression(in(beta(cells(RPL14,RPLP2,RPS5,FOXA2,RPL22);RPL11(@>(Validated(targets(of(C@MYC(transcriptional(activation(TAF4B,TERT,TP53,MYC,CREBBP)0.00122045 RPL5,RPS7(@>(NRAS(1.39e@03);RPS19,RPL35A,RPS7(@>(MYC(2.25e@04);RPS19(@>(IRAK1(3.08e@02);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(TRIM28(1.64e@02);RPS26,RPS7(@>(HAUS8(1.64e@03);(@>(CCNE1(6.83e@03);RPS26,RPS7(@>(FTSJ2(1.78e@02);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(NUDT1(2.14e@03);RPS26,RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7(@>(RPS5(3.05e@05);RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A(@>(RPLP2(3.96e@04);RPS19,RPS10(@>(POLD4(2.33e@02);RPS26,RPS19,RPS7(@>(TP53(1.72e@03);RPS26,RPS19,RPS7(@>(TACC3(8.31e@04);RPS26,RPS19,RPS7(@>(GEMIN4(2.74e@02);RPS26,RPS10,RPL35A(@>(RNMTL1(3.46e@04);RPS10,RPL35A(@>(ZNF497(1.11e@03);RPS26,RPS19,RPS10,RPL35A,RPS7(@>(TXNL4A(3.23e@02);RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7(@>(RPL22(3.92e@05);RPS10(@>(C9orf142(3.77e@03);(@>(TERT(1.23e@02);RPS19(@>(TRAF4(3.89e@02);RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7(@>(RPL14(2.97e@04);RPL11,RPL35A,RPS7(@>(RBMX(2.78e@02)0 RPL11(@>(FHIT;RPL35A(@>(MECOM;RPL5(@>(RPLP2;RPS10(@>(RPL14;RPS17(@>(RPS5;RPS19(@>(WWOX,FHIT;RPS24(@>(MECOM;RPS26(@>(RPLP2;RPS7(@>(RPL140.20041667 RPL11(@>(RPL14;RPL5(@>(TP53;RPS10(@>(ESR1;RPS17(@>(RPL22;RPS19(@>(RPS5;RPS24(@>(MYC,RPL14;RPS26(@>(ESR1;RPS7(@>(RPLP2Inherited(Anomalies(of(the(Skin UCEC 1 TERT 3.29E@08 ATP2A2(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,CREBBP,PIK3R1,PIK3CA);TERT(@>(IL2(signaling(events(mediated(by(PI3K(PIK3CA,MYC,TERT,PIK3R1);TERT(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(MYC,TERT,TP53,MZF1,ESR1);TERT(@>(telomeres(telomerase(cellular(aging(and(immortality(MYC,TERT,TP53,RB1,KRAS);KRT1,TERT(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CTNNB1,CCND1,MYC,TERT);TERT(@>(role(of(nicotinic(acetylcholine(receptors(in(the(regulation(of(apoptosis(PIK3CA,TERT,PIK3R1);TERT(@>(Validated(targets(of(C@MYC(transcriptional(activation(TAF4B,TERT,TP53,MYC,CREBBP);TINF2,TERT,DKC1(@>(Regulation(of(Telomerase(CCND1,MYC,TERT,SIN3A,ESR1)0.183228108 1 1Combined(Heart(and(Skeletal(Defects UCEC 0.63 CREBBP 2.76E@30 CREBBP(@>(nfat(and(hypertrophy(of(the(heart((NFATC1,CREBBP,PIK3R1,PIK3CA);EP300,CREBBP(@>(IFN@gamma(pathway(PIK3CA,CREBBP,PIK3R1);EP300,CREBBP(@>(Direct(p53(effectors(VDR,TP53,PIDD,RB1,PTEN,CREBBP);EP300,CREBBP(@>(FOXA1(transcription(factor(network(FOXA2,CREBBP,NKX3@1,ESR1);EP300,CREBBP(@>(transcription(regulation(by(methyltransferase(of(carm1(CREBBP,PRKAR1B);EP300(@>(cell(cycle:(g2/m(checkpoint(TP53,MYT1);EP300,CREBBP(@>(carm1(and(regulation(of(the(estrogen(receptor(CREBBP,ESR1);CREBBP(@>(wnt(signaling(pathway(CTNNB1,CCND1,MYC,CREBBP);EP300(@>(Validated(nuclear(estrogen(receptor(alpha(network(CCND1,MYC,UBE2M,ESR1);EP300(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CTNNB1,CCND1,MYC,TERT);EP300,CREBBP(@>(mechanism(of(gene(regulation(by(peroxisome(proliferators(via(ppara(MYC,PRKAR1B,RB1,CREBBP);CREBBP(@>(regulation(of(transcriptional(activity(by(pml(TP53,CREBBP,RB1);EP300,CREBBP(@>(E2F(transcription(factor(network(MYC,CCNE1,TRIM28,CREBBP,RB1);EP300,CREBBP(@>(il@7(signal(transduc�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 0.33848148 1Hereditary(Sensory(Neuropathy UCEC 0.71 NEFL,DNM2 0.096035489 NTRK1(@>(Trk(receptor(signaling(mediated(by(PI3K(and(PLC@gamma(CCND1,PIK3CA,NRAS,PIK3R1,KRAS);NDRG1(@>(Direct(p53(effectors(VDR,TP53,PIDD,RB1,PTEN,CREBBP);NTRK1(@>(trka(receptor(signaling(pathway(PIK3CA,PIK3R1);NTRK1(@>(p73(transcription(factor(network(MYC,RNF43,RB1,WWOX);NTRK1(@>(p75(NTR)@mediated(signaling(PIK3CA,IRAK1,OMG,TP53,PIK3R1);DNM2(@>(PAR1@mediated(thrombin(signaling(events(PIK3CA,GNAQ,PIK3R1,DNM2);PMP22(@>(a6b1(and(a6b4(Integrin(signaling(ERBB2,PIK3CA,PIK3R1,ERBB3)0.130354569 1 1Li(Fraumeni(and(Related(Syndromes UCEC 0.63 TP53 1.96E@18 TP53(@>(chaperones(modulate(interferon(signaling(pathway(TP53,RB1);TP53(@>(Direct(p53(effectors(VDR,TP53,PIDD,RB1,PTEN,CREBBP);TP53(@>(LKB1(signaling(events(MYC,TP53,ESR1);TP53,CHEK2(@>(cell(cycle:(g2/m(checkpoint(TP53,MYT1);TP53(@>(estrogen(responsive(protein(efp(controls(cell(cycle(and(breast(tumors(growth(TP53,ESR1);TP53(@>(overview(of(telomerase(protein(component(gene(htert(transcriptional(regulation(MYC,TERT,TP53,MZF1,ESR1);CDKN2A(@>(Coregulation(of(Androgen(receptor(activity(CTNNB1,CCND1,CASP8,NKX3@1);CDKN2A,TP53(@>(Hypoxic(and(oxygen(homeostasis(regulation(of(HIF@1@alpha(TP53,ARNT);TP53(@>(telomeres(telomerase(cellular(aging(and(immortality(MYC,TERT,TP53,RB1,KRAS);CDKN2A(@>(Regulation(of(nuclear(beta(catenin(signaling(and(target(gene(transcription(CTNNB1,CCND1,MYC,TERT);TP53(@>(btg(family(proteins(and(cell(cycle(regulation(CCND1,TP53,RB1);TP53(@>(Transcriptional((activation(of((cell(cycle(inhibitor(p21(TP53);TP53,CHEK2(@>(PLK3(signaling(events(CCNE1,TP53);TP53(@>(p53(signaling(pathway(CCND1,CCNE1,TP53,�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.458169296 0.57574242 1Lipoprotein(Deficiencies UCEC 1 1 0.064216795 MTTP,APOB,LCAT,SAR1B,APOA1(@>(A1BG(1.19e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(SLC27A5(7.36e@03);MTTP,APOB,LCAT,SAR1B,APOA1(@>(HPD(1.19e@02);MTTP,APOB,LCAT,SAR1B,APOA1(@>(ATRN(2.26e@02);MTTP,APOB,LCAT,SAR1B,APOA1,ABCA1(@>(NEU4(1.39e@02)1 1

Page 133: Rachel D. Melamed

122

 Supplementary  Table  4:  Continuation  of  Supplementary  Table  3  

     

MD C gene_enrichmentgeneIntersection pathway_correlationpathway coex_CG coexpression humannet_sethumannet biogrid_set biogrid !Chronic!Granulomatous!Disease LUAD 1 0.384190056 0.078679242 NCF4,CYBA!E>!CCND3(3.54eE02);NCF2,NCF4,CYBA!E>!ZGPAT(1.74eE02);NCF2,NCF4,CYBB,CYBA!E>!RIT1(1.48eE02);NCF2,NCF4,CYBB,CYBA!E>!ITGAX(1.79eE02);NCF2,NCF4,CYBB,CYBA!E>!DNAJC5(1.82eE02);NCF2,NCF4,CYBB,CYBA!E>!PTGER4(1.60eE02);NCF2,NCF4,CYBB,CYBA!E>!SIRPB1(1.30eE02);NCF2,NCF4,CYBB,CYBA!E>!TNFSF13B(4.88eE02);NCF2,NCF4,CYBB,CYBA!E>!PQLC1(1.33eE02);NCF2,NCF4,CYBB,CYBA!E>!TBL1X(1.72eE02);NCF4,CYBA!E>!ARHGEF6(2.95eE02);NCF2,NCF4,CYBB,CYBA!E>!GNG2(1.37eE02);NCF2,NCF4,CYBB,CYBA!E>!NFATC1(1.81eE02);NCF2,NCF4!E>!U2AF1(3.61eE02);NCF2,NCF4,CYBB,CYBA!E>!BTK(1.94eE02);NCF2,NCF4,CYBB,CYBA!E>!SAMSN1(1.35eE02);NCF2,NCF4,CYBB,CYBA!E>!LPAR6(2.19eE02);NCF2,NCF4!E>!IL18RAP(1.83eE02);NCF2,NCF4,CYBB,CYBA!E>!RB1(3.72eE02);NCF2,NCF4,CYBB!E>!ADNP2(3.36eE02);NCF2,NCF4,CYBB,CYBA!E>!MFSD7(3.07eE02);NCF2,NCF4,CYBB,CYBA!E>!PMAIP1(1.26eE02);NCF4,CYBB,CYBA!E>!TBX21(1.81eE02);NCF2,NCF4,CYBB,CYBA!E>!ANKRD44(1.82eE02);NCF2,NCF4,CYBB,CYBA!E>!AQP9(1.28eE02);NCF2,NCF4,CYBB,CYBA!E>!PPM1F(1.53eE02);NCF2,NCF4,CYBB,CYBA!E>!CTDP1(1.37eE02);NCF2,NCF4,CYBB,CYB���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Congenital!Ichthyosis LUAD 1 1 0.020062495 ALOX12B,SPINK5,CSTA,TGM1!E>!SPRR3(8.05eE03);ALOX12B!E>!KRT28(3.04eE03);!E>!FLG(1.68eE03);ALOX12B,SPINK5,CSTA,ABCA12,TGM1!E>!SERPINB13(1.68eE03);ALOX12B,SPINK5,KRT2,ABCA12!E>!POF1B(8.05eE03)1 1 !Disorders!of!Phosphorous!Metabolism LUAD 1 CYP27B1 0.003900738 FGF23!E>!SyndecanE2Emediated!signaling!events(HRAS,LAMA3,FGF19,NF1);CYP27B1!E>!Vitamin!D!(calciferol)!metabolism(GC,CYP27B1);FGF23!E>!SyndecanE1Emediated!signaling!events(COL11A1,COL5A2,MET,FGF19,COL5A1,COL3A1)1 1 1 !DiamondEBlackfan!Anemia LUAD 1 1 0.006070507 RPS26,RPS19,RPS10,RPS7!E>!BYSL(4.40eE02);RPL5,RPS10,RPL11!E>!CCND3(3.53eE02);!E>!UTF1(8.74eE03);RPS24,RPL5!E>!ALG10(3.42eE02);RPS26,RPS19,RPS7!E>!GEMIN4(2.58eE02);RPS26!E>!CDKN2A(9.01eE03);RPS26,RPL35A,RPS7!E>!U2AF1(4.63eE04);RPS26,RPS7!E>!FANCD2(1.47eE02);RPS26,RPS19,RPS10,RPL35A!E>!LAGE3(4.10eE02);!E>!CTCFL(1.14eE02);!E>!MDM2(4.11eE02);!E>!VENTX(7.84eE03);RPS26,RPS19,RPS7!E>!TP53(1.67eE03);RPS26,RPS19,RPS10,RPS7!E>!EIF4EBP1(1.12eE02);RPS26,RPS19,RPS7!E>!TFDP1(2.13eE03);RPS26,RPS19,RPS10,RPL35A,RPS7!E>!TXNL4A(3.14eE02);!E>!TERT(1.10eE02);RPS19,RPS7!E>!CNIH1(4.86eE04);RPL5,RPS7!E>!NRAS(1.31eE03);RPS26,RPS10,RPL35A!E>!RNMTL1(3.54eE04);RPS26,RPS7!E>!METTL1(2.41eE02);RPS26,RPS19,RPS10,RPS7!E>!HAX1(1.49eE02)1 1 !Inherited!Anomalies!of!the!Skin LUAD 1 TERT 3.07EE07 TERT!E>!overview!of!telomerase!protein!component!gene!htert!transcriptional!regulation(TERT,TP53);TERT!E>!telomeres!telomerase!cellular!aging!and!immortality(TERT,TP53,RB1,KRAS);KRT1,TERT!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CDKN2A,TBL1X,CCND1,APC,TERT,SMARCA4)0.086097896 KRT6A,KRT16!E>!CYP27B1(1.45eE02);WRAP53,TERC,DKC1,NHP2,NOP10!E>!GEMIN4(4.63eE02);TERC,NHP2,NOP10!E>!FAM58A(4.53eE02);KRT6A,TERC,NHP2,KRT16,NOP10!E>!HRAS(4.57eE02);TERC,NHP2,NOP10!E>!SLC10A3(4.68eE02);KRT6A,NHP2,KRT16,NOP10!E>!NXN(4.37eE02);!E>!AKR1B10(4.64eE02)1 1 !Spinocerebellar!Ataxia LUAD 1 ATM 8.60EE07 PRKCG!E>!EGFR!Inhibitor!Pathway,!Pharmacodynamics(ERBB2,NRAS,HRAS,EGFR,KRAS);ATM!E>!apoptotic!signaling!in!response!to!dna!damage(ATM,TP53);ATM!E>!p53!pathway(CDKN2A,ATM,TP53,MDM2);ATM!E>!ATM!pathway(MDM2,ATM,FANCD2);ATM!E>!cell!cycle:!g2/m!checkpoint(MDM2,ATM,TP53,MYT1);ATM!E>!ATM!mediated!response!to!DNA!doubleEstrand!break(ATM);ATM!E>!cdc25!and!chk1!regulatory!pathway!in!response!to!dna!damage(ATM,MYT1);ATM!E>!role!of!brca1!brca2!and!atr!in!cancer!susceptibility(ATM,TP53,FANCD2);ATM!E>!BARD1!signaling!events(ATM,TP53,FANCD2);ATM!E>!atm!signaling!pathway(MDM2,ATM,TP53);PRKCG!E>!IL8E!and!CXCR2Emediated!signaling!events(PLCB1,GNG2,HCK);ATM!E>!rb!tumor!suppressor/checkpoint!signaling!in!response!to!dna!damage(ATM,TP53,RB1,MYT1);PRKCG!E>!IL8E!and!CXCR1Emediated!signaling!events(PLCB1,GNG2,HCK);ATM!E>!Metformin!Pathway,!Pharmacodynamic(ATM,STK11,PRKAA1);ATM!E>!E2F!transcription!factor!network(TFDP1,CCND3,ATM,CDKN2A,RB1);ATM!E>!hypoxia!and!p53!in!the!cardiovascular!system(MDM2,ATM,TP53);TBP!E>!Glucocorticoid!rece����������������������������������������������������������������������������������������������������������������������������������������������������0.001274155 SYT14!E>!SCG2(2.09eE02);APTX,ZNF592,ATXN2,TTBK2,TBP,KCNC3,ITPR1,NOP56,SETX,SYT14,C10orf2!E>!KIAA0907(2.58eE03);JPH3,TDP1,KCNC3,FGF14!E>!SAMD10(1.63eE02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!DOC2B(3.57eE03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!C1orf173(3.57eE03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!TMEM132D(4.61eE03);JPH3,APTX,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!NF1(2.40eE03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!OPCML(2.50eE03);JPH3,ZNF592,SPTBN2,TTBK2,KCNC3,PRNP,PRKCG,FGF14,PPP2R2B,ATXN1!E>!DNAJC5(3.21eE02);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B!E>!TTC33(2.02eE03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,TBP,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !GlucoseE6EPhosphate!Dehydrogenase!DeficiencyLUAD 0.05 UBL4A,G6PD 0.034828736 NA!E>!NA(NA) E1 1 1 !Hypopituitarism LUAD 1 BTK 0.034828736 FGFR1!E>!SyndecanE2Emediated!signaling!events(HRAS,LAMA3,FGF19,NF1);BTK!E>!bcr!signaling!pathway(NFATC1,HRAS,BTK,PPP3CA);BTK!E>!EPO!signaling!pathway(HRAS,PTPN11,BTK);POU1F1!E>!Glucocorticoid!receptor!regulatory!network(NFATC1,TP53,TBX21,MDM2,SMARCA4,GATA3);GH1!E>!trefoil!factors!initiate!mucosal!healing(ERBB2,HRAS,EGFR);FGFR1!E>!SyndecanE1Emediated!signaling!events(COL11A1,COL5A2,MET,FGF19,COL5A1,COL3A1)1 0.2886 1 !Combined!Heart!and!Skeletal!Defects LUAD 1 3.78EE08 EP300,CREBBP!E>!Direct!p53!effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);EP300,CREBBP!E>!p53!pathway(CDKN2A,ATM,TP53,MDM2);EP300!E>!cell!cycle:!g2/m!checkpoint(MDM2,ATM,TP53,MYT1);EP300!E>!Validated!transcriptional!targets!of!AP1!family!members!Fra1!and!Fra2(CCND1,NFATC1,LAMA3,CDKN2A);EP300!E>!Validated!transcriptional!targets!of!TAp63!isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);EP300!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CDKN2A,TBL1X,CCND1,APC,TERT,SMARCA4);CREBBP!E>!regulation!of!transcriptional!activity!by!pml(TP53,RB1);EP300,CREBBP!E>!E2F!transcription!factor!network(TFDP1,CCND3,ATM,CDKN2A,RB1);EP300!E>!p73!transcription!factor!network(JAG2,MDM2,NTRK1,RB1,WWOX);EP300!E>!hypoxia!and!p53!in!the!cardiovascular!system(MDM2,ATM,TP53);EP300,CREBBP!E>!Glucocorticoid!receptor!regulatory!network(NFATC1,TP53,TBX21,MDM2,SMARCA4,GATA3);EP300!E>!melanocyte!development!and!pigmentation!pathway(HRAS,KIT)0.71670305 1 1 !Neurofibromatosis LUAD 0.66 NF1 0.001098828 NF1!E>!Regulation!of!Ras!family!activation(HRAS,NRAS,NF1,KRAS);NF1!E>!SyndecanE2Emediated!signaling!events(HRAS,LAMA3,FGF19,NF1);NF1!E>!chromatin!remodeling!by!hswi/snf!atpEdependent!complexes(ARID1A,SMARCA4,NF1)0.675271975 0.22904762 NF2!E>!CDKN2A 0.60125 !Hereditary!Sensory!Neuropathy LUAD 1 NTRK1 1.95EE05 NTRK1!E>!Trk!receptor!signaling!mediated!by!PI3K!and!PLCEgamma(CCND1,HRAS,NRAS,NTRK1,KRAS);NTRK1!E>!ARMSEmediated!activation(NTRK1,BRAF);NDRG1!E>!Direct!p53!effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);EGR2!E>!Validated!transcriptional!targets!of!TAp63!isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);NTRK1!E>!TRKA!activation!by!NGF(NTRK1);NTRK1!E>!NGF!signalling!via!TRKA!from!the!plasma!membrane(NTRK1);NTRK1!E>!Signalling!to!ERKs(NTRK1);NTRK1!E>!Signalling!to!STAT3(NTRK1);NTRK1!E>!trka!receptor!signaling!pathway(HRAS,NTRK1);RAB7A!E>!IL8E!and!CXCR2Emediated!signaling!events(PLCB1,GNG2,HCK);NTRK1!E>!Frs2Emediated!activation(NTRK1,BRAF);NTRK1!E>!Signalling!to!p38!via!RIT!and!RIN(NTRK1,BRAF);NTRK1!E>!p73!transcription!factor!network(JAG2,MDM2,NTRK1,RB1,WWOX);PMP22!E>!a6b1!and!a6b4!Integrin!signaling(ERBB2,MET,HRAS,LAMA3,EGFR)0.128961218 1 1 !Severe!Combined!Immunodeficiency LUAD 1 0.009297242 DCLRE1C!E>!ATM!pathway(MDM2,ATM,FANCD2);ADA!E>!Validated!transcriptional!targets!of!TAp63!isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);ADA!E>!p73!transcription!factor!network(JAG2,MDM2,NTRK1,RB1,WWOX);JAK3,IL2RG!E>!IL2Emediated!signaling!events(HRAS,NRAS,PTPN11,KRAS);JAK3!E>!il!6!signaling!pathway(PTPN11,HRAS);ADA!E>!Validated!transcriptional!targets!of!deltaNp63!isoforms(COL5A1,CDKN2A,ATM,MDM2)0.001684618 CIITA,RFX5,DCLRE1C!E>!C11orf35(2.10eE02);!E>!GATA3(2.09eE02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1!E>!CCND3(1.40eE04);ZAP70,IL2RG,JAK3,ADA,PTPRC,IL7R,CD3D!E>!ZGPAT(2.99eE03);ZAP70,IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D!E>!PTGER4(1.27eE02);IL2RG,CIITA,JAK3,ADA,PNP,PTPRC!E>!TNFSF13B(4.44eE02);RFXANK!E>!RBM10(2.42eE03);ZAP70,IL2RG,RFXAP,RFX5,AK2,IL7R,CD3D,DCLRE1C!E>!ALG10(7.50eE04);IL2RG,RFX5,DCLRE1C!E>!CMTR2(2.45eE03);IL2RG,JAK3,PNP,RFX5,PTPRC,DCLRE1C!E>!AOAH(2.05eE02);IL2RG,JAK3,PTPRC!E>!TBL1X(2.39eE02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!ARHGEF6(9.58eE04);ZAP70,IL2RG,CIITA,JAK3,RFXAP,RFX5,PTPRC,CD3D,DCLRE1C!E>!ARID1A(5.42eE03);ZAP70,IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,DCLRE1C!E>!GNG2(3.39eE02);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC!E>!NFATC1(1.94eE02);ADA,PNP,AK2!E>!U2AF1(9.03eE04);ZAP70,IL2RG,JAK3,RFXAP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1!E>!ARID2(1.03eE02);IL2RG,CIITA,JAK3,RFX5,PTPRC,DCLRE1C!E>!BTK(3.08eE03);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1!E>!THEMIS(2.35eE04);RFXA��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Specified!Hamartoses LUAD 0.97 STK11 2.59EE05 VHL!E>!vegf!hypoxia!and!angiogenesis(HRAS,KDR,ARNT);PTEN!E>!Direct!p53!effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);VHL!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53,ARNT);STK11!E>!Metformin!Pathway,!Pharmacodynamic(ATM,STK11,PRKAA1)0.717635623 0.25727907 1 !Li!Fraumeni!and!Related!Syndromes LUAD 0.09 CDKN2A,TP53 4.67EE28 TP53!E>!chaperones!modulate!interferon!signaling!pathway(TP53,RB1);TP53!E>!apoptotic!signaling!in!response!to!dna!damage(ATM,TP53);TP53!E>!Direct!p53!effectors(BCL2L14,PMAIP1,TP53,EGFR,RB1,MET,APC,MDM2,SMARCA4,TFDP1);TP53,CDKN2A,CHEK2!E>!p53!pathway(CDKN2A,ATM,TP53,MDM2);CHEK2!E>!ATM!pathway(MDM2,ATM,FANCD2);TP53,CHEK2!E>!cell!cycle:!g2/m!checkpoint(MDM2,ATM,TP53,MYT1);CDKN2A!E>!Validated!transcriptional!targets!of!AP1!family!members!Fra1!and!Fra2(CCND1,NFATC1,LAMA3,CDKN2A);CDKN2A!E>!Validated!transcriptional!targets!of!TAp63!isoforms(MDM2,CDKN2A,PMAIP1,SPATA18);TP53!E>!overview!of!telomerase!protein!component!gene!htert!transcriptional!regulation(TERT,TP53);TP53,CHEK2!E>!role!of!brca1!brca2!and!atr!in!cancer!susceptibility(ATM,TP53,FANCD2);CDKN2A,TP53!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53,ARNT);TP53!E>!telomeres!telomerase!cellular!aging!and!immortality(TERT,TP53,RB1,KRAS);CDKN2A!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CDKN2A,TBL1X,CCND1�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.719639888 0 CDKN2A!E>!MDM2;CHEK2!E>!CCND3,ATM;TP53!E>!TP53,SMAD41 !Genetic!Anomalies!of!Leukocytes LUAD 1 0.02688935 ITGB2!E>!Beta2!integrin!cell!surface!interactions(ITGAX,SPON2,FGB)0.471117151 1 1 !Lipoprotein!Deficiencies LUAD 1 MTTP 1 0.097427692 APOB,LCAT,SAR1B,APOA1!E>!SLC26A1(2.83eE02);APOB,LCAT,SAR1B,APOA1!E>!PCK1(3.92eE02);APOB,LCAT,SAR1B,APOA1!E>!PROS1(1.82eE02);!E>!AKR1C2(2.10eE02);APOB,LCAT,SAR1B,APOA1!E>!EHHADH(1.83eE02);APOB,LCAT,SAR1B,APOA1!E>!BHMT(1.95eE02);APOB,LCAT,SAR1B,APOA1!E>!GBA3(1.93eE02);APOB,LCAT,SAR1B,APOA1!E>!ABCG5(2.40eE02);APOB,LCAT,SAR1B,APOA1!E>!MTTP(2.00eE02);!E>!CD5L(2.00eE02)1 1 !Disorders!of!Urea!Cycle!Metabolism LUAD 1 1 0.069332922 ASS1,NAGS,ARG1,ASL,CPS1!E>!GC(1.50eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!SLC26A1(1.99eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!FGB(2.95eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!PROS1(1.02eE02);ASS1,NAGS,ASL,CPS1!E>!HSBP1L1(9.88eE03);ASS1!E>!AKR1C2(3.55eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!EHHADH(1.20eE02);NAGS,ARG1,ASL,CPS1!E>!SOWAHB(1.38eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!CYP4V2(3.55eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!GBA3(8.52eE03);ASS1,NAGS,ARG1,ASL,CPS1!E>!ABCG5(9.51eE03);ASS1,NAGS,ARG1,ASL,CPS1!E>!MTTP(9.88eE03);!E>!CD5L(9.22eE03)1 1 !Retinitis!Pigmentosa LUAD 1 PDE6B 0.828420314 6.10EE15 RPGR,CRX,SNRNP200,CA4,EYS,CRB1,CERKL,PRPF3,TULP1,C2orf71,TOPORS,FAM161A!E>!KIAA0907(3.68eE02);KLHL7,SPATA7,CRB1,MERTK,CERKL,FAM161A!E>!DOC2B(3.41eE02);!E>!MUC16(2.78eE04);SNRNP200,CRB1,PRPF31,CERKL,IDH3B,PRPF8,FAM161A!E>!UCKL1(2.83eE03);CRX,FSCN2,RDH12,PRPH2,CNGB1,EYS,CRB1,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3!E>!GPR112(5.36eE07);KLHL7,SPATA7,CRB1,PRCD,IMPG2,C2orf71,FAM161A!E>!TMEM132D(3.43eE02);SNRNP200!E>!FZD10(1.68eE02);IMPDH1!E>!CYP27B1(1.33eE02);CRX,SNRNP200,KLHL7,EYS,SPATA7,CRB1,CERKL,USH2A,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!GTF2I(4.98eE02);!E>!SLC22A6(1.68eE02);CNGA1!E>!ANKRD37(1.67eE02);CRX,FSCN2,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,MERTK,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!RIMS2(3.67eE04);SPATA7,CRB1,PRPF3!E>!ITGB8(1.07eE02);RPGR,CA4,IMPDH1,PRPF3,BEST1,RP2,TOPORS,SEMA4A!E>!TBL1X(2.76eE02);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,EYS,CRB1,NRL,PDE6A,PDE6G,CERKL,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,IMPG2,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4!E>!RP1L1(1.27eE17);CRX,FSCN�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.02531579 CNGA1!E>!CNGA2,LRRC32,EIF4G3;CNGB1!E>!PABPC3,PDE6B,PDE6B,TXNL4A;IMPDH1!E>!PRPF6,PABPC3;PDE6A!E>!TXNL4A;PDE6G!E>!PRPF6,EIF4G3;PRPF3!E>!PABPC3;PRPF31!E>!TXNL4A;PRPF8!E>!PRPF6,U2AF1,U2AF1;RP9!E>!PABPC3;SNRNP200!E>!TXNL4A1 !Haemophilia LUAD 0.7 F8 0.071266219 VWF!E>!Platelet!Aggregation!Inhibitor!Pathway,!Pharmacodynamics(COL3A1,COL4A2,FGB,PLCB1);F8,F9!E>!intrinsic!prothrombin!activation!pathway(F8,PROS1,COL4A2,FGB)1 0.7955 0.52680952 !Chronic!Granulomatous!Disease LUSC 1 1 0.067971546 NCF2,NCF4,CYBB,CYBA!E>!B2M(1.31eE02);NCF2,NCF4,CYBB,CYBA!E>!MARCH1(8.59eE03);NCF2,NCF4,CYBB!E>!USP25(1.12eE02);NCF2,NCF4,CYBB,CYBA!E>!NFE2L2(8.87eE03);NCF2,NCF4,CYBB,CYBA!E>!LYZ(2.21eE02);NCF2,NCF4,CYBB,CYBA!E>!PTEN(1.06eE02);NCF2,NCF4,CYBB,CYBA!E>!HDAC10(1.05eE02);NCF2,NCF4,CYBB,CYBA!E>!KDM5A(1.96eE02);NCF2,NCF4,CYBB,CYBA!E>!NFATC1(1.20eE02);NCF4,CYBB,CYBA!E>!CHKB(2.32eE02);NCF2,NCF4,CYBB,CYBA!E>!TRABD(1.01eE02);NCF2,NCF4,CYBB,CYBA!E>!CREBBP(1.06eE02);NCF2,NCF4,CYBB,CYBA!E>!ODF3B(1.41eE02);NCF2,NCF4,CYBB,CYBA!E>!CTDP1(9.25eE03);NCF2,NCF4,CYBB,CYBA!E>!REL(9.24eE03);NCF2,NCF4,CYBB,CYBA!E>!KDM6A(1.20eE02);NCF2,NCF4,CYBB,CYBA!E>!METRNL(9.24eE03);NCF2,NCF4,CYBB,CYBA!E>!PQLC1(8.59eE03);NCF2,NCF4,CYBB,CYBA!E>!LPAR6(1.21eE02);NCF2,NCF4,CYBB,CYBA!E>!BID(1.21eE02);NCF2,NCF4,CYBB,CYBA!E>!RB1(1.92eE02);NCF2,NCF4,CYBB,CYBA!E>!PIM3(8.92eE03);NCF4!E>!NINJ2(1.01eE02);NCF2,NCF4,CYBB,CYBA!E>!EVI2A(1.71eE02);NCF2,NCF4,CYBB,CYBA!E>!EVI2B(9.25eE03);NCF2,NCF4!E>!EXOC3(2.80eE02);NCF2,NCF4,CYBB,CYBA!E>!NOTCH1(8.21eE03);NCF2,NCF4,CY�����������������������������������������������������������1 1 !Congenital!Ichthyosis LUSC 1 1 0.006200982 ALOX12B,ALOXE3,SPINK5,CSTA,KRT2,ABCA12,TGM1!E>!CERS3(3.87eE04);CSTA,NIPAL4,LIPN,ABHD5!E>!NFE2L2(3.14eE02);ALOX12B,SPINK5,CSTA,TGM1!E>!PAX9(3.14eE02);ALOX12B,ALOXE3,CSTA,TGM1!E>!PARD6G(4.70eE02);CSTA,NIPAL4,LIPN,ABHD5!E>!NOTCH1(2.92eE02);ABCA12,TGM1!E>!EGFR(3.98eE02)0.79744737 1 !DiamondEBlackfan!Anemia LUSC 1 0.959794083 0.002784462 RPS26!E>!CDKN2A(9.26eE03);!E>!YEATS4(1.30eE03);RPS26,RPS19,RPS7!E>!TP53(1.51eE03);RPS26,RPS19,RPL35A,RPS7!E>!PDCD6(1.33eE04);RPS26,RPS19,RPS10,RPL35A,RPS7!E>!TXNL4A(3.08eE02);!E>!NMU(2.86eE02);RPL5,RPS7!E>!TTF2(4.93eE02);RPS26,RPS19,RPS7!E>!TYMS(2.02eE03);RPS26,RPS7!E>!TRIP13(9.59eE03)1 1 !Spinocerebellar!Ataxia LUSC 1 0.977629639 0.01980021 JPH3,CACNA1A,POLG,ATXN2,SYNE1,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,PPP2R2B!E>!SBF1(9.84eE03);JPH3,ZNF592,ATM,TTBK2,KCNC3,ITPR1,SETX,PPP2R2B!E>!MARCH1(3.18eE02);JPH3,CACNA1A,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B!E>!L1CAM(1.18eE02);JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1!E>!CCSER1(2.45eE03);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2B!E>!PARK2(2.80eE03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!MAPK8IP2(2.90eE03);TDP1,ATM,TBP,NOP56,SYT14,C10orf2!E>!CCDC77(1.54eE02);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!CLCN4(2.99eE03);POLG,ATXN2,TBP,NOP56,AFG3L2,C10orf2!E>!BRD9(2.95eE03);ZNF592,POLG,ATXN7,ATXN2,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,ATXN1!E>!BRD1(3.18eE02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!CSMD3(2.99eE03);JPH3,CACNA1A��������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 0.73607576 !Combined!Heart!and!Skeletal!Defects LUSC 0.6 CREBBP 3.46EE14 CREBBP!E>!nfat!and!hypertrophy!of!the!heart!(NFATC1,CREBBP,PIK3CA);CREBBP!E>!inhibition!of!huntingtons!disease!neurodegeneration!by!histone!deacetylase!inhibitors(CREBBP);EP300,CREBBP!E>!Direct!p53!effectors(TP53,BID,EGFR,RB1,PTEN,CREBBP);EP300,CREBBP!E>!p53!pathway(CDKN2A,TP53,CREBBP);CREBBP!E>!NotchEHLH!transcription!pathway(CREBBP);EP300,CREBBP!E>!mechanism!of!gene!regulation!by!peroxisome!proliferators!via!ppara(FAT1,RB1,CREBBP);CREBBP!E>!regulation!of!transcriptional!activity!by!pml(TP53,CREBBP,RB1);EP300,CREBBP!E>!E2F!transcription!factor!network(TYMS,CDKN2A,CREBBP,RB1);EP300,CREBBP!E>!ilE7!signal!transduction(PIK3CA,CREBBP);EP300!E>!p73!transcription!factor!network(MAPK11,CDK6,RB1,WWOX);EP300,CREBBP!E>!Glucocorticoid!receptor!regulatory!network(NFATC1,TP53,MAPK11,CREBBP);CREBBP!E>!Signaling!events!mediated!by!Stem!cell!factor!receptor!(cEKit)(PIK3CA,CREBBP,PTEN);EP300!E>!ATFE2!transcription!factor!network(MAPK11,RB1,NF1);CREBBP!E>!Signaling!events!mediated!by!TCPTP(PIK3CA,EGFR,CREBBP);EP300,CREBBP!E>!F����������������������������������������������������1 0.0925 CREBBP!E>!TP53,TP53;EP300!E>!CREBBP1 !Neurofibromatosis LUSC 0.53 NF1 0.236599603 0.719411434 0 NF1!E>!EVI2A,CDKN2A;NF2!E>!NF11 !Severe!Combined!Immunodeficiency LUSC 1 0.23074305 0.001138137 ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!B2M(6.47eE03);IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,DCLRE1C!E>!PLEKHO1(2.51eE02);ZAP70,IL2RG,CIITA,JAK3,PTPRC,IL7R,CD3D,DCLRE1C!E>!BRD1(1.02eE02);!E>!YEATS4(1.46eE02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!HDAC10(2.23eE04);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!KDM5A(8.24eE03);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC!E>!NFATC1(1.79eE02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!FOXP1(2.39eE02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!CHKB(1.16eE03);ZAP70,IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!TRABD(1.29eE04);RFXAP,RFX5,DCLRE1C!E>!PRDM15(8.47eE03);ZAP70,IL2RG,JAK3,PNP,RFX5,PTPRC,IL7R,DCLRE1C!E>!CREBBP(3.99eE02);ZAP70,IL2RG,CIITA,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!CTDP1(3.26eE03);IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC!E>!REL(3.66eE02);ZAP70,IL2RG,CIITA,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!KDM6A(2.09eE02);ZAP70,IL2RG,RFX5,DCLRE1C!E>!ZBED4(2.61eE05);ADA,AK2,DCLRE1C!E>!CDK6(2.85eE04);ZA�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Specified!Hamartoses LUSC 0.63 PTEN 8.39EE06 PTEN!E>!skeletal!muscle!hypertrophy!is!regulated!via!aktEmtor!pathway(PIK3CA,PTEN);PTEN!E>!regulation!of!eifE4e!and!p70s6!kinase(PIK3CA,PTEN);PTEN!E>!mtor!signaling!pathway(PTEN,PIK3CA);PTEN!E>!Direct!p53!effectors(TP53,BID,EGFR,RB1,PTEN,CREBBP);VHL!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53);PTEN!E>!RhoA!signaling!pathway(PTEN,MAPK12,SLC9A3);PTEN!E>!pten!dependent!cell!cycle!arrest!and!apoptosis(PTEN,PIK3CA);PTEN!E>!Signaling!events!mediated!by!Stem!cell!factor!receptor!(cEKit)(PIK3CA,CREBBP,PTEN)0.619582187 0 PTEN!E>!TTF2;SDHB!E>!EGFR;SDHD!E>!CDKN2A;STK11!E>!TP531 !Li!Fraumeni!and!Related!Syndromes LUSC 0.03 CDKN2A,TP53 5.79EE15 TP53!E>!Fluoropyrimidine!Pathway,!Pharmacodynamics(TYMS,TP53);TP53!E>!chaperones!modulate!interferon!signaling!pathway(TP53,RB1);TP53!E>!apoptotic!signaling!in!response!to!dna!damage(BID,TP53);TP53!E>!Direct!p53!effectors(TP53,BID,EGFR,RB1,PTEN,CREBBP);TP53,CDKN2A,CHEK2!E>!p53!pathway(CDKN2A,TP53,CREBBP);CDKN2A!E>!CEMYC!pathway(CDKN2A,FBXW7);TP53!E>!estrogen!responsive!protein!efp!controls!cell!cycle!and!breast!tumors!growth(TP53,CDK6);CDKN2A,TP53!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53);TP53!E>!telomeres!telomerase!cellular!aging!and!immortality(TP53,RB1);TP53!E>!btg!family!proteins!and!cell!cycle!regulation(TP53,RB1);TP53!E>!Transcriptional!!activation!of!!cell!cycle!inhibitor!p21(TP53);TP53!E>!p53!signaling!pathway(TP53,RB1);TP53!E>!regulation!of!transcriptional!activity!by!pml(TP53,CREBBP,RB1);TP53!E>!rb!tumor!suppressor/checkpoint!signaling!in!response!to!dna!damage(TP53,RB1);CDKN2A!E>!E2F!transcription!factor!network(TYMS,CDKN2A,CREBBP,RB1);TP53!E>!Glucocorticoid!receptor�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.651962919 0.02405 CDKN2A!E>!PTEN;CHEK2!E>!CDK6,TP53,PTEN;TP53!E>!CREBBP1 !Lipoprotein!Deficiencies LUSC 1 0.885354678 0.097427692 APOB,LCAT,APOA1!E>!ENOSF1(1.81eE02);MTTP,APOB,LCAT,SAR1B,APOA1!E>!PROS1(1.81eE02);!E>!AKR1C2(2.16eE02);APOB,LCAT,APOA1!E>!SLC6A12(2.16eE02);MTTP,APOB,LCAT,APOA1!E>!SELO(1.87eE02)1 1 !Retinitis!Pigmentosa LUSC 1 EYS 1 5.69EE14 KLHL7,SPATA7,CRB1,MERTK,PDE6B,CERKL,PRCD,FAM161A!E>!CLCN4(4.22eE02);TTC8,CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,CRB1,NRL,PDE6B,PDE6G,CERKL,GUCA1B,TULP1,NR2E3,ROM1,USH2A,RP1,PRCD,IMPG2,C2orf71,SAG,RBP3,FAM161A!E>!EYS(3.56eE16);TTC8!E>!COLEC12(1.19eE03);TTC8,CRX,FSCN2,PRPH2,CNGB1,KLHL7,SPATA7,CRB1,MERTK,PDE6B,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!KCNIP4(1.35eE03);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4!E>!CLUL1(2.31eE14);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4!E>!SLC6A13(1.01eE12);TTC8,CRX,LRAT,FSCN2,RDH12,PRPH2,CNGB1,SPATA7,CRB1,MERTK,CERKL,TULP1,ROM1,PROM1,CNGA1,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!UNC13B(4.65eE06);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,RP1,PRCD,RPE65,SAG,RBP3,ABCA4!E>!KCNJ13(4.97eE12)1 1 !Haemophilia LUSC 1 0.00999936 F8,F9!E>!intrinsic!prothrombin!activation!pathway(COL4A5,PROS1)1 1 1 !Chronic!Granulomatous!Disease PRAD 1 1 0.078898011 NCF2,NCF4,CYBA!E>!GPS2(1.49eE02);NCF2,NCF4,CYBB,CYBA!E>!SPOPL(1.38eE02);NCF2,NCF4,CYBB,CYBA!E>!COTL1(3.89eE02);!E>!CRISPLD2(4.13eE02);NCF2,NCF4,CYBB,CYBA!E>!TMUB2(1.62eE02);NCF2!E>!MFI2(2.93eE02);CYBA!E>!ARRDC1(2.54eE02);NCF2,NCF4,CYBB,CYBA!E>!HELZ2(1.23eE02);NCF2,NCF4,CYBB,CYBA!E>!PTEN(1.32eE02);NCF2,NCF4,CYBB!E>!PAK2(1.35eE02);NCF2,NCF4,CYBB,CYBA!E>!NFATC1(1.34eE02);NCF2,NCF4,CYBB,CYBA!E>!CRTC2(1.42eE02);NCF2,NCF4,CYBB,CYBA!E>!ADRBK1(1.57eE02);NCF2,NCF4,CYBA!E>!ZGPAT(1.40eE02);NCF2,NCF4!E>!EGR3(2.34eE02);NCF2,NCF4,CYBA!E>!GPR160(1.26eE02);NCF2,CYBB,CYBA!E>!HNMT(1.21eE02);NCF2,NCF4,CYBB!E>!SENP5(3.88eE02);NCF2,NCF4,CYBB,CYBA!E>!DNAJC5(1.35eE02);NCF2,NCF4,CYBB,CYBA!E>!ANKRD13D(1.27eE02);NCF2,NCF4,CYBB,CYBA!E>!CTDP1(1.30eE02);NCF4,CYBA!E>!CDKN1B(3.93eE02);CYBA!E>!POLD4(2.93eE02);CYBA!E>!RPS27(2.87eE02);NCF2,NCF4,CYBB,CYBA!E>!TPD52L2(1.12eE02);NCF2,NCF4,CYBA!E>!GMEB2(1.40eE02);NCF2,NCF4,CYBB,CYBA!E>!PQLC1(1.16eE02);NCF2,NCF4,CYBB,CYBA!E>!RGS19(2.46eE02);NCF2,NCF4,CYBB,CYBA!E>!TOR4A(1.38eE02);NCF2,NCF4,CYBB!E>!R�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Glycogenosis PRAD 1 1 0.078728665 PGAM2,PHKB,PHKA2,AGL!E>!SLC25A30(1.28eE02)1 1 !Congenital!Ichthyosis PRAD 1 1 0.02654965 ALOX12B,SPINK5,CSTA,TGM1!E>!C9orf169(2.49eE03);ALOX12B,ALOXE3,CSTA,TGM1!E>!PARD6G(3.82eE02);ALOXE3,CSTA,ABCA12,TGM1!E>!ARRDC1(2.44eE02);NIPAL4,KRT2,LIPN,ABCA12!E>!NRARP(4.60eE03);SPINK5,CSTA,NIPAL4,ABCA12,TGM1!E>!PTK6(4.60eE03)1 1 !Disorders!of!Phosphorous!Metabolism PRAD 0.69 SLC34A3 0.000406927 SLC34A3!E>!Type!II!Na+/Pi!cotransporters(SLC34A3) 1 1 1 !Spinocerebellar!Ataxia PRAD 1 1 0.027197561 APTX,ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,SETX,SYT14,C10orf2,PPP2R2B!E>!CASP8AP2(3.80eE03);JPH3,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1!E>!CREBL2(1.97eE02);JPH3,APTX,CACNA1A,ATXN2,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B!E>!DLG1(5.25eE03);JPH3,TDP1,KCNC3,FGF14!E>!SAMD10(1.32eE02);JPH3,CACNA1A,ATXN2,PDYN,SPTBN2,TTBK2,KCNC3,PRKCG,FGF14,SYT14,PPP2R2B!E>!NSMF(9.90eE03);JPH3,ZNF592,CACNA1A,ATXN7,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,FGF14,SETX,PPP2R2B,ATXN1!E>!PHC3(2.61eE03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!GRIN1(3.68eE03);JPH3,CACNA1A,ATXN2,SYNE1,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!MYT1(2.84eE03);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,ATXN10,PRKCG,FGF14,SYT14,PPP2R2B!E>!ZNF285(1.11eE02);JPH3,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN10,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!STMN3(2.84eE03);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ATXN������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Specified!Hamartoses PRAD 0.66 PTEN 0.062379972 STK11!E>!Metformin!Pathway,!Pharmacodynamic(SLC2A4,CRTC2);PTEN!E>!RhoA!signaling!pathway(CDKN1B,PTEN);PTEN!E>!pten!dependent!cell!cycle!arrest!and!apoptosis(CDKN1B,PTEN);PTEN!E>!Negative!regulation!of!the!PI3K/AKT!network(PTEN)0.57662662 1 1 !Retinitis!Pigmentosa PRAD 1 1 6.10EE15 CRX,EYS,SPATA7,CRB1,MERTK,PDE6B,CERKL,PRCD,IMPG2,C2orf71,RBP3!E>!CREBL2(4.56eE02);CRX,FSCN2,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,PDE6B,CERKL,TULP1,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!MYT1(7.66eE06);SNRNP200,CRB1,PRPF31,CERKL,IDH3B,PRPF8,FAM161A!E>!UCKL1(1.68eE03);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,EYS,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4!E>!KCNG4(2.10eE17);CRX,LRAT,FSCN2,RDH12,PRPH2,RHO,CNGB1,EYS,CRB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,BEST1,PROM1,CNGA1,RP1,PRCD,C2orf71,RPE65,SAG,RBP3,FAM161A,ABCA4!E>!SAMD7(1.45eE17);RPGR,CA4,PRPF3,BEST1,RP2,TOPORS,SEMA4A!E>!GPR160(1.80eE02);SNRNP200,EYS!E>!THSD7B(1.89eE05);!E>!WFDC1(1.49eE05);!E>!NXPH2(5.64eE03);TTC8,RDH12,SPATA7,MERTK,CNGA1,FAM161A!E>!ZFHX3(1.68eE03);CRX,FSCN2,RDH12,SNRNP200,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,PDE6B,CERKL,PRPF3,TULP1,ROM1,PRCD,IMPG2,C2orf71,TOPORS,RBP3,FAM161A!E>!PCMTD2(7.66eE06);ZNF513,KLHL7,SPATA7,CRB1,PDE6B,CERKL,FAM161A!E>!TMEM145(2.�������0.34981818 0.26722222 !Long!QT!Syndrome READ 0.69 CACNA1C 3.14EE21 CACNA1C!E>!Nicotine!Pathway!(Chromaffin!Cell),!Pharmacodynamics(CACNA1C);CACNA1C!E>!Sympathetic!Nerve!Pathway!(PreE!and!PostE!Ganglionic!Junction)(CACNA1C,TH);CACNA1C!E>!AntiEdiabetic!Drug!Potassium!Channel!Inhibitors!Pathway,!Pharmacodynamics(PDX1,CACNA1C,INS)1 1 1 !Chronic!Granulomatous!Disease READ 1 1 0.07714767 !E>!CRLF2(3.23eE02);NCF2,NCF4,CYBB,CYBA!E>!PRPF3(2.17eE02);NCF2,NCF4,CYBB,CYBA!E>!CACNA2D4(1.29eE02);NCF2!E>!KRAS(4.79eE02);NCF4,CYBA!E>!TRAPPC2(2.54eE02);NCF2,NCF4,CYBB,CYBA!E>!PLAGL2(1.33eE02);NCF2,NCF4,CYBB,CYBA!E>!CSF2RA(1.45eE02);NCF2,CYBB,CYBA!E>!TCF7L2(1.37eE02);NCF2,NCF4,CYBA!E>!TCEANC(2.36eE02);NCF2,NCF4,CYBB!E>!CTNNBL1(1.42eE02);NCF2,NCF4!E>!RNF40(3.72eE02);NCF2,NCF4,CYBB,CYBA!E>!IRF2(1.53eE02);NCF2,NCF4,CYBB,CYBA!E>!CR1(1.22eE02);NCF2,NCF4,CYBB,CYBA!E>!FBRS(1.23eE02);NCF2,CYBB,CYBA!E>!HS3ST3B1(1.16eE02);NCF4!E>!C17orf103(2.44eE02);NCF2,NCF4,CYBB,CYBA!E>!IL3RA(1.69eE02);!E>!ADK(3.11eE02);NCF2,CYBB!E>!RAB9A(1.57eE02);NCF2,CYBB!E>!EMP1(4.04eE02);NCF2!E>!SRCAP(3.53eE02);NCF2,NCF4,CYBB,CYBA!E>!RAB39A(1.60eE02);NCF2,NCF4,CYBB,CYBA!E>!MAP2K3(2.17eE02);NCF2,NCF4,CYBB,CYBA!E>!MOSPD2(1.16eE02);NCF2,NCF4,CYBB,CYBA!E>!MCL1(1.19eE02);NCF2!E>!P2RY8(1.27eE02)1 1 !Glycogenosis READ 0.86 PHKG2 6.77EE12 AGL,PYGM,PHKA1,PHKB,PHKA2,PHKG2!E>!Glycogen!breakdown!(glycogenolysis)(PHKG2,GYG2)0.372442654 1 0.36782353 !Inherited!Anomalies!of!the!Skin READ 1 0.004631074 TERT!E>!HIFE1Ealpha!transcription!factor!network(SMAD4,MCL1,SMAD3);TERT!E>!telomeres!telomerase!cellular!aging!and!immortality(TP53,KRAS);TERT!E>!Validated!targets!of!CEMYC!transcriptional!activation(SMAD4,TP53,SMAD3)0.10458489 1 1 !Spinocerebellar!Ataxia READ 1 ATXN10 0.008075087 CACNA1A!E>!Sympathetic!Nerve!Pathway!(PreE!and!PostE!Ganglionic!Junction)(CACNA1C,TH);PRNP!E>!Glypican!1!network(HCK,FGFR1);CACNA1A!E>!AntiEdiabetic!Drug!Potassium!Channel!Inhibitors!Pathway,!Pharmacodynamics(PDX1,CACNA1C,INS);TBP!E>!Validated!targets!of!CEMYC!transcriptional!repression(ERBB2,SMAD4,SMAD3)0.013136049 ATXN7,TDP1,SYNE1,ATM,TTBK2,TBP,ITPR1,PPP2R2B!E>!FHIT(3.07eE03);!E>!PPP2R3B(4.28eE02);JPH3,ATXN2,TTBK2,TBP,PPP2R2B!E>!CDRT4(4.76eE02);JPH3,APTX,CACNA1A,ATXN2,PDYN,SYNE1,SPTBN2,TTBK2,KCNC3,PRNP,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B!E>!GPM6B(3.44eE03);ZNF592,POLG,ATXN7,TDP1,ATM,TBP,ITPR1,SETX,ATXN1!E>!PRPF3(8.48eE03);JPH3,ZNF592,CACNA1A,ATXN2,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1!E>!CCSER1(2.41eE03);JPH3,ZNF592,CACNA1A,ATXN7,ATXN2,PDYN,TDP1,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,SETX,SYT14,PPP2R2B,ATXN1!E>!WHSC1L1(1.11eE03);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,PRKCG,FGF14,AFG3L2,SYT14,PPP2R2B!E>!PARK2(2.44eE03);JPH3,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,PPP2R2B!E>!ENSA(4.21eE02);JPH3,ZNF592,CACNA1A,ATXN2,PDYN,SYNE1,ATM,SPTBN2,TTBK2,KCNC3,ITPR1,PRKCG,FGF14,SYT14,PPP2R2B,ATXN1!E>!ZNF785(1.51eE03);JPH3,ZNF592,CACNA1A,ATXN7,SYNE1,ATM,SPTBN2,TTBK2,TBP,KCNC3,ITPR1,FGF14,SETX,SYT14,PPP2R2B,ATXN1!E>!CUL5(1.42eE03);POLG,ATXN7,TDP1,SYNE1,ATM,TBP,ITPR1,�����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Severe!Combined!Immunodeficiency READ 1 1 0.017157155 IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,DCLRE1C!E>!PRPF3(1.29eE02);ZAP70,IL2RG,CIITA,JAK3,RFXAP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!OFD1(1.32eE03);RFXAP,DCLRE1C!E>!FNTA(4.56eE02);PNP!E>!CACNA2D4(4.19eE02);ZAP70,IL2RG,JAK3,RFXAP,IL7R,CD3D,DCLRE1C!E>!CSTF2T(1.10eE02);!E>!DDX47(3.75eE02);IL2RG,JAK3,RFXANK,RFX5,PTPRC!E>!PRR14(6.34eE03);IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,DCLRE1C!E>!PLAGL2(1.07eE02);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!TCEANC(2.91eE03);ZAP70,IL2RG,JAK3,RFXAP,RFX5,IL7R,CD3D,DCLRE1C!E>!ASXL1(1.37eE03);PNP!E>!CTNNBL1(3.18eE02);IL2RG,JAK3,ADA,PNP,PTPRC!E>!HCK(4.98eE02);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,DCLRE1C!E>!IRF2(9.30eE03);IL2RG,JAK3,ADA,PNP,PTPRC!E>!HS3ST3B1(1.89eE02);JAK3,PTPRC!E>!C17orf103(4.06eE02);RFX5!E>!GPRC5D(1.63eE02);CIITA!E>!FLT3(1.41eE03);RFXAP,AK2,DCLRE1C!E>!FANCB(2.62eE02);CIITA,RFX5,DCLRE1C!E>!ADK(1.46eE02);PNP,PTPRC!E>!P2RY8(2.19eE03);RFXAP!E>!ANK1(4.68eE02)1 1 !Lipoprotein!Deficiencies READ 1 1 0.075339054 MTTP,APOB,LCAT,SAR1B,APOA1!E>!KLC4(3.60eE02);APOB,LCAT,APOA1!E>!FCN3(2.07eE02);MTTP!E>!INSEIGF2(2.31eE02);APOB,ABCA1!E>!HS3ST3B1(2.70eE02);!E>!ADK(3.79eE02);MTTP,APOB,LCAT,APOA1!E>!ARSD(1.18eE02);MTTP,APOB,LCAT,SAR1B,APOA1!E>!ARSE(2.07eE02)1 1 !Disorders!of!Urea!Cycle!Metabolism READ 1 1 0.088390959 ASS1,NAGS,ARG1,ASL,CPS1!E>!KLC4(4.33eE02);NAGS,ARG1,ASL,CPS1!E>!FCN3(2.80eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!ARSD(1.57eE02);ASS1,NAGS,ARG1,ASL,CPS1!E>!ARSE(2.80eE02)1 1 !Retinitis!Pigmentosa READ 1 PRPF3 1 4.04EE13 !E>!IMMP2L(2.15eE05);!E>!DHRSX(1.34eE03);EYS,CERKL!E>!NKX6E3(1.42eE02);RPGR,CRX,FSCN2,RDH12,PRPH2,CNGB1,EYS,CRB1,TULP1,ROM1,RP2,PRCD,IMPG2,C2orf71,RBP3,SEMA4A!E>!CACNA2D4(1.12eE08);FSCN2,PRPH2,RHO,CNGB1,PDE6A,PDE6G,GUCA1B,RLBP1,RGR,NR2E3,ROM1,CNGA1,RP1,SAG,ABCA4!E>!EGFL6(4.38eE09);CRX,FSCN2,RDH12,PRPH2,CNGB1,EYS,CRB1,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3!E>!ASMT(1.12eE08);TTC8,CRX,FSCN2,RDH12,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,MERTK,PDE6B,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!KIAA1467(6.02eE04);!E>!ZBED1(3.70eE03);TTC8,CRX,FSCN2,PRPH2,CNGB1,KLHL7,EYS,SPATA7,CRB1,TULP1,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!GSG1(6.99eE07);CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6G,GUCA1B,TULP1,NR2E3,ROM1,PROM1,CNGA1,RP1,PRCD,C2orf71,SAG,RBP3,FAM161A,ABCA4!E>!KERA(5.05eE15);FAM161A!E>!SHOX(9.94eE11)1 0.40083333 !DopaEResponsive!Dystonia READ 0.62 TH 2.34EE32 TH!E>!AlphaEsynuclein!signaling(HCK,TH,PARK2);TH!E>!Sympathetic!Nerve!Pathway!(PreE!and!PostE!Ganglionic!Junction)(CACNA1C,TH)1 1 1 !Congenital!Ectodermal!Dysplasia SKCM 1 0.28548436 0.000106457 !E>!FAM58A(9.62eE03);LAMB3,ITGB4,PLEC,KRT5,LAMC2!E>!ANAPC15(2.97eE03);!E>!EIF3D(1.58eE03);ITGA6,LAMB3,ITGB4,GJB6,COL17A1,PLEC,COL7A1,KRT5,KRT14,LAMC2,LAMA3!E>!PTK6(2.00eE06);LAMB3,ITGB4,KRT5!E>!MYC(2.50eE04);!E>!CDKN2A(2.84eE02);PLEC!E>!PHGDH(3.27eE02);PLEC!E>!RPL13(4.57eE03);!E>!CCND1(2.48eE02);LAMB3,ITGB4,PLEC,KRT5!E>!PPDPF(2.47eE03);!E>!LSM12(2.57eE02);!E>!SLC25A39(1.42eE02);!E>!MRP63(4.21eE02);!E>!SRMS(2.88eE02);ITGA6,LAMB3,ITGB4,COL17A1,KRT5,KRT14,LAMC2,LAMA3!E>!TDRP(5.01eE04);PLEC!E>!GRN(1.79eE02);!E>!TP53(1.56eE02);!E>!TSPAN31(1.19eE02);!E>!HDAC3(3.26eE03);ITGB4,PLEC,KRT5!E>!ACD(5.04eE03);!E>!KRT78(1.13eE02);GJB6!E>!TCHHL1(9.83eE03);PLEC!E>!DEF8(2.72eE02);ITGA6,COL17A1,COL7A1,KRT14!E>!DSG1(1.08eE05);LAMB3,ITGB4,PLEC,KRT5,KRT14,LAMC2!E>!TNFRSF6B(6.12eE04);!E>!RPTN(3.38eE04);PLEC!E>!TPD52L2(4.98eE02);LAMB3,ITGB4,PLEC,COL7A1,KRT5!E>!CHMP1A(1.44eE02);!E>!DYNAP(9.82eE03);ITGB4,PLEC,COL7A1!E>!SLC2A4RG(6.70eE03);!E>!FOLR3(2.43eE03)1 1 !Chronic!Granulomatous!Disease SKCM 1 1 0.078763609 NCF2,NCF4,CYBB,CYBA!E>!ZFX(1.54eE02);NCF2,NCF4,CYBB,CYBA!E>!HELZ2(1.79eE02);NCF2,NCF4!E>!KIAA1257(2.33eE02);NCF2!E>!STK19(3.34eE02);NCF2,NCF4,CYBB,CYBA!E>!B2M(2.08eE02);NCF2,NCF4,CYBB!E>!VPS9D1(1.84eE02);NCF2,NCF4,CYBB,CYBA!E>!DDX3X(1.31eE02);NCF2,NCF4,CYBB,CYBA!E>!TMUB2(1.76eE02);NCF2,NCF4,CYBB,CYBA!E>!CLEC2B(2.70eE02);NCF2,NCF4,CYBB,CYBA!E>!ADAM8(1.31eE02);NCF2,NCF4,CYBB,CYBA!E>!PTEN(1.49eE02);NCF2,NCF4,CYBB,CYBA!E>!ZNF276(1.43eE02);NCF4!E>!SLC25A39(1.56eE02);NCF2,NCF4,CYBB,CYBA!E>!GNAI2(1.51eE02);NCF2,NCF4,CYBB,CYBA!E>!PPP6C(1.49eE02);NCF2,NCF4,CYBB,CYBA!E>!RPGRIP1(1.45eE02);NCF4,CYBB!E>!POM121(3.52eE02);NCF2,NCF4,CYBA!E>!ZGPAT(1.56eE02);!E>!SERPINB10(1.35eE02);NCF2,NCF4,CYBB,CYBA!E>!GRN(3.32eE02);NCF2,NCF4,CYBB,CYBA!E>!DNAJC5(1.52eE02);NCF2,NCF4,CYBB,CYBA!E>!RBM22(1.44eE02);NCF4!E>!ITGA2B(3.33eE02);CYBB,CYBA!E>!OXA1L(1.33eE02);NCF2,NCF4,CYBB!E>!MPP7(1.54eE02);!E>!SLC4A1(3.39eE02);NCF2,NCF4,CYBB,CYBA!E>!CLEC5A(1.39eE02);NCF2,NCF4,CYBB,CYBA!E>!GPR141(1.49eE02);NCF4,CYBB,CYBA!E>!ITGA4(1.49eE02);NCF2,NCF4,CYB������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Congenital!Ichthyosis SKCM 1 1 0.005136028 SPINK5,CSTA,NIPAL4,ABCA12,TGM1!E>!PTK6(3.02eE03);SPINK5,CSTA,NIPAL4,KRT2,LIPN,ABHD5!E>!MPP7(1.32eE02);ALOX12B,SPINK5,CSTA,TGM1!E>!KRT78(5.11eE03);ALOX12B,ABCA12!E>!TCHHL1(3.36eE04);ALOX12B,SPINK5,CSTA,ABCA12,TGM1!E>!DSG1(2.89eE04);ALOX12B,SPINK5,CSTA,ABCA12,TGM1!E>!RPTN(2.89eE04);ALOX12B,SPINK5,CSTA,TGM1!E>!DYNAP(3.02eE03)1 1 !Polycystic!Kidney,!Autosomal!Dominant SKCM 1 0.013304666 TSC2!E>!LKB1!signaling!events(RPTOR,MYC,TP53);TSC2!E>!Validated!targets!of!CEMYC!transcriptional!repression(CCND1,HDAC3,EP300,MYC)0.736679678 1 1 !Inherited!Anomalies!of!the!Skin SKCM 1 TERT 5.99EE19 TERT!E>!erk1/erk2!mapk!signaling!pathway(MYC,TERT);TERT!E>!IL2!signaling!events!mediated!by!PI3K(MYC,TERT,RAC1);TERT!E>!overview!of!telomerase!protein!component!gene!htert!transcriptional!regulation(MYC,TERT,TP53);TERT!E>!telomeres!telomerase!cellular!aging!and!immortality(MYC,TERT,TP53);KRT1,TERT!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);TERT!E>!role!of!nicotinic!acetylcholine!receptors!in!the!regulation!of!apoptosis(TERT,FASLG);TERT!E>!Validated!targets!of!CEMYC!transcriptional!activation(TERT,TP53,UBTF,CDK4,MYC,EP300);TINF2,TERT,DKC1!E>!Regulation!of!Telomerase(CCND1,MYC,TERT,SAP18,ACD)0.164716031 1 1 !Combined!Heart!and!Skeletal!Defects SKCM 0.64 EP300 2.01EE19 EP300!E>!cell!cycle:!g2/m!checkpoint(EP300,TP53,MYT1);EP300!E>!Validated!transcriptional!targets!of!AP1!family!members!Fra1!and!Fra2(CCND1,EP300,CDKN2A);EP300,CREBBP!E>!acetylation!and!deacetylation!of!rela!in!nucleus(HDAC3,EP300);EP300!E>!Notch!signaling!pathway(CCND1,MYC,EP300,NOTCH2,DNER);EP300!E>!Validated!nuclear!estrogen!receptor!alpha!network(CCND1,KLRC3,EP300,MYC);EP300!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);EP300!E>!Validated!targets!of!CEMYC!transcriptional!repression(CCND1,HDAC3,EP300,MYC);EP300,CREBBP!E>!E2F!transcription!factor!network(CCNE2,EP300,MYC,CDKN2A);EP300,CREBBP!E>!ilE7!signal!transduction(ITGA2B,EP300);EP300,CREBBP!E>!Validated!targets!of!CEMYC!transcriptional!activation(TERT,TP53,UBTF,CDK4,MYC,EP300);EP300!E>!hypoxia!and!p53!in!the!cardiovascular!system(EP300,TP53);EP300!E>!ATFE2!transcription!factor!network(CCND1,EP300,CDK4,NF1);EP300!E>!melanocyte!development!and!pigmentation!pathway(MITF,EP300);EP300,CREBBP!E>����������������������������������������������������������������1 0.32066667 1 !Specified!Anomalies!of!the!Musculoskeletal!SystemSKCM 1 MITF 0.097075171 MITF,SNAI2!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);MITF!E>!IL6Emediated!signaling!events(MITF,MYC,RAC1);MITF!E>!melanocyte!development!and!pigmentation!pathway(MITF,EP300)1 1 1 !Neurofibromatosis SKCM 0.61 NF1 4.53EE05 NF1!E>!Regulation!of!Ras!family!activation(NRAS,RASA2,NF1);NF1!E>!ATFE2!transcription!factor!network(CCND1,EP300,CDK4,NF1)0.719411434 0.222 NF2!E>!CDKN2A 1 !Tuberous!Sclerosis SKCM 1 0.013304666 TSC2,TSC1!E>!LKB1!signaling!events(RPTOR,MYC,TP53);TSC2!E>!Validated!targets!of!CEMYC!transcriptional!repression(CCND1,HDAC3,EP300,MYC)0.594661126 1 1 !Severe!Combined!Immunodeficiency SKCM 1 1 0.001386824 IL2RG,JAK3,ADA,PNP,PTPRC!E>!HELZ2(4.66eE02);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C,RAG2,RAG1!E>!THEMIS(1.76eE04);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!B2M(6.45eE03);NHEJ1,RFX5,AK2!E>!EIF3D(4.81eE05);ZAP70,IL2RG,JAK3,ADA,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!KCNAB2(3.23eE03);ZAP70,RFXANK,IL7R,CD3D!E>!RPL13(2.74eE03);AK2!E>!TERT(1.68eE02);ZAP70,IL2RG,JAK3,PTPRC,IL7R,CD3D,DCLRE1C!E>!TC2N(5.55eE03);ZAP70,IL2RG,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!CLEC2B(3.84eE03);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!CLEC2D(1.81eE04);IL2RG,JAK3,PNP,RFX5,PTPRC!E>!ADAM8(3.98eE02);ZAP70,IL2RG,CIITA,JAK3,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!ZNF276(3.22eE03);ZAP70,IL2RG,JAK3,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!PCED1B(1.76eE03);JAK3!E>!SLC25A39(7.97eE04);IL2RG,CIITA,PNP,RFX5,PTPRC,DCLRE1C!E>!LY86(1.75eE02);ZAP70,IL2RG,JAK3,PNP,RFX5,PTPRC,DCLRE1C!E>!PPP6C(5.65eE03);ZAP70,IL2RG,JAK3,RFXAP,PNP,RFX5,PTPRC,IL7R,CD3D,DCLRE1C!E>!POM121(8.56eE04);ZAP70,IL2RG,JAK3,ADA,PTPRC,IL7R,CD3D!E>!ZGPAT(2.74eE03);PNP,AK2!E>!����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !Specified!Hamartoses SKCM 0.72 PTEN 0.061006438 STK11!E>!LKB1!signaling!events(RPTOR,MYC,TP53);VHL!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53);STK11!E>!Regulation!of!AMPK!activity!via!LKB1(RPTOR);PTEN!E>!pten!dependent!cell!cycle!arrest!and!apoptosis(PTEN,FASLG)0.561087039 0.13528125 PTEN!E>!TP53;SDHB!E>!CDKN2A;STK11!E>!NRAS,OGDHL,TP530.2405 PTEN!E>!TCEB3C;SDHB!E>!TP53;STK11!E>!CCNE2;VHL!E>!CHGB!Li!Fraumeni!and!Related!Syndromes SKCM 0.05 CDKN2A,TP53 2.63EE31 TP53!E>!LKB1!signaling!events(RPTOR,MYC,TP53);TP53,CHEK2!E>!cell!cycle:!g2/m!checkpoint(EP300,TP53,MYT1);CDKN2A!E>!Validated!transcriptional!targets!of!AP1!family!members!Fra1!and!Fra2(CCND1,EP300,CDKN2A);TP53!E>!estrogen!responsive!protein!efp!controls!cell!cycle!and!breast!tumors!growth(CDK4,TP53);TP53!E>!overview!of!telomerase!protein!component!gene!htert!transcriptional!regulation(MYC,TERT,TP53);TP53,CHEK2!E>!role!of!brca1!brca2!and!atr!in!cancer!susceptibility(TP53,FANCA);CDKN2A,TP53!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53);TP53!E>!telomeres!telomerase!cellular!aging!and!immortality(MYC,TERT,TP53);CDKN2A!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(TERT,CCND1,CDKN2A,MITF,MYC,EP300);TP53!E>!btg!family!proteins!and!cell!cycle!regulation(CCND1,TP53);TP53!E>!Transcriptional!!activation!of!!cell!cycle!inhibitor!p21(TP53);TP53!E>!p53!signaling!pathway(CCND1,CDK4,TP53);TP53!E>!rb!tumor!suppressor/checkpoint!signaling!in!response!to!dna!damage(CDK4����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.662892321 0.15561765 CDKN2A!E>!PTEN;CHEK2!E>!CDK4,TP53,PTEN;TP53!E>!EP3001 !Lipoprotein!Deficiencies SKCM 1 1 0.058815075 APOB,LCAT,APOA1!E>!SPTLC3(3.68eE02);APOB,SAR1B!E>!IDH1(2.34eE02);MTTP,APOB,LCAT,SAR1B,APOA1!E>!UGT2B15(6.62eE03);MTTP,APOB,LCAT,SAR1B,APOA1,ABCA1!E>!C2(1.13eE02);APOB,APOA1!E>!TIMD4(3.68eE02);APOB,APOA1!E>!STAB2(1.22eE02);MTTP,APOB,LCAT,SAR1B,APOA1!E>!APCS(1.13eE02)1 1 !Retinitis!Pigmentosa SKCM 1 EYS,CRB1 1 2.00EE13 CERKL,IDH3B!E>!ASB16(5.43eE03);CRX,FSCN2,KLHL7,SPATA7,MERTK,PDE6B,CERKL,ROM1,PRCD,C2orf71,RBP3,FAM161A!E>!SV2B(3.21eE03);CRX,FSCN2,PRPH2,CNGB1,KLHL7,SPATA7,PDE6B,CERKL,TULP1,PRCD,IMPG2,C2orf71,RBP3,FAM161A!E>!MYT1(4.84eE05);TTC8,CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,KLHL7,SPATA7,NRL,MERTK,PDE6B,PDE6G,CERKL,GUCA1B,TULP1,NR2E3,ROM1,PRCD,IMPG2,C2orf71,SAG,RBP3,FAM161A,ABCA4!E>!CRB1(2.61eE05);CRX,FSCN2,RDH12,PRPH2,CNGB1,CERKL,TULP1,ROM1,PRCD,IMPG2,C2orf71,RBP3!E>!LRTM1(5.49eE09);TTC8,CRX,FSCN2,RDH12,PRPH2,CNGB1,SPATA7,PDE6B,TULP1,ROM1,PRCD,RBP3!E>!FMN1(2.31eE04);CRX,FSCN2,RDH12,PRPH2,RHO,CNGB1,NRL,PDE6B,PDE6A,PDE6G,GUCA1B,RLBP1,TULP1,RGR,NR2E3,ROM1,PROM1,CNGA1,RP1,PRCD,C2orf71,SAG,RBP3,FAM161A,ABCA4!E>!PROL1(6.23eE14);KLHL7!E>!NUDT4(2.32eE02);KLHL7,IDH3B!E>!SCN5A(4.86eE02);LRAT,SNRNP200,PROM1!E>!UTF1(4.41eE02);ZNF513,KLHL7,SPATA7,MERTK,PDE6B,PRCD,FAM161A!E>!CHRNA4(1.59eE02);CRX,SNRNP200,CA4,CERKL,TULP1,PROM1,FAM161A!E>!KCNB2(9.21eE04);TTC8,RPGR,SPATA7,CERKL,FAM161A!E>!CHGB(4.86eE02);!E>!MITF(1.67eE04);!E>!SPTBN5(4.63eE�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.21645 IDH3B!E>!IDH1,EIF2B1;IMPDH1!E>!SKIV2L;PROM1!E>!DDX3X;PRPF3!E>!RPL13,CRB1,PRPF6,PRPF6,DDX3X;PRPF31!E>!PRPF6,RPGRIP1,MYC;PRPF8!E>!PRPF60.22446667 CA4!E>!SLC4A1,EP300;CRX!E>!RBFOX1;FAM161A!E>!PARK2,RUNDC3A,MYC;IDH3B!E>!TP53,DDX3X,HDAC3,PRPF6,FANCA;IMPDH1!E>!PRPF6,DDX3X;NR2E3!E>!ITGA4;PRPF3!E>!HDAC5;PRPF31!E>!MYC;PRPF8!E>!PRPF6,PHGDH,RPGRIP1,HDAC5;ROM1!E>!ITGA4;RPGR!E>!PARK2;SNRNP200!E>!MYC;TOPORS!E>!PRPF6,UPF3A!Hereditary!Hemorrhagic!Telangiectasia SKCM 1 1.86EE06 SMAD4!E>!LKB1!signaling!events(RPTOR,MYC,TP53);SMAD4!E>!Validated!nuclear!estrogen!receptor!alpha!network(CCND1,KLRC3,EP300,MYC);SMAD4!E>!Validated!targets!of!CEMYC!transcriptional!repression(CCND1,HDAC3,EP300,MYC);SMAD4!E>!Validated!targets!of!CEMYC!transcriptional!activation(TERT,TP53,UBTF,CDK4,MYC,EP300)1 1 1 !Disorders!of!Aromatic!Amino!Acid!MetabolismSKCM 1 MC1R 1 0.072585714 BLOC1S6,AP3B1,BLOC1S3,HPS5,HPS1!E>!HELZ2(1.61eE02);BLOC1S6,AP3B1,BLOC1S3,HPS5,HPS6,HPS1!E>!TMUB2(3.76eE02);TYR,OCA2,TYRP1,SLC45A2,BLOC1S6,AP3B1,DTNBP1,BLOC1S3,HPS5,HPS1!E>!KCNAB2(9.33eE03);BLOC1S6,AP3B1,BLOC1S3,HPS6,HPS1!E>!GNAI2(2.33eE02);BLOC1S6,BLOC1S3,HPS5,HPS6,HPS1!E>!GRN(1.01eE02);!E>!OXA1L(2.24eE02);BLOC1S6,AP3B1,HPS5!E>!CLEC5A(3.63eE02);BLOC1S6,BLOC1S3,HPS5,HPS6,HPS1!E>!TPD52L2(3.55eE02);TAT,HPD,FAH!E>!C2(3.91eE02);HPD!E>!LIME1(4.55eE02);BLOC1S6,AP3B1,BLOC1S3,HPS5,HPS1!E>!RGS19(3.70eE02);TYR,OCA2,TYRP1,SLC45A2!E>!MITF(2.04eE02);BLOC1S3,HPS6,HPS1!E>!INPPL1(2.36eE02);TYR,OCA2,TYRP1,SLC45A2,BLOC1S3,HPS5,HPS1!E>!MAD1L1(1.01eE02);BLOC1S6,DTNBP1,BLOC1S3,HPS1!E>!TCF25(3.81eE02);BLOC1S6,AP3B1,DTNBP1,BLOC1S3,HPS5,HPS1!E>!SBNO2(2.33eE02);HPS1!E>!FOLR2(1.98eE02)1 1 !Chronic!Granulomatous!Disease STAD 1 0.846838069 0.072585714 NCF2,NCF4,CYBB,CYBA!E>!B2M(2.04eE02);NCF2,NCF4,CYBB,CYBA!E>!DIAPH2(1.14eE02);NCF2,NCF4,CYBB,CYBA!E>!TRPS1(9.36eE03);NCF2!E>!KRAS(4.90eE02);CYBA!E>!CLECL1(3.26eE02);NCF2,NCF4,CYBB,CYBA!E>!CLEC2B(2.77eE02);NCF2,NCF4,CYBB,CYBA!E>!PTEN(1.14eE02);NCF2,NCF4,CYBB,CYBA!E>!SNX2(1.75eE02);NCF2,NCF4,CYBB,CYBA!E>!CLEC12A(1.60eE02);NCF4!E>!IL5RA(3.13eE02);NCF2,CYBB,CYBA!E>!RHOA(9.22eE03);NCF2,NCF4,CYBB,CYBA!E>!IRF2(9.69eE03);!E>!KLRF1(3.81eE02);!E>!CLEC1B(3.30eE02);NCF4,CYBA!E>!CD69(3.41eE02);NCF2,NCF4,CYBB,CYBA!E>!DYRK1A(2.15eE02);NCF2,CYBB,CYBA!E>!UAP1L1(1.35eE02);!E>!PLGRKT(3.41eE02);CYBB,CYBA!E>!DPP7(1.97eE02);NCF2,CYBB,CYBA!E>!CD44(9.47eE03);NCF2,NCF4,CYBB,CYBA!E>!CD274(1.06eE02)1 1 !Disorders!of!Phosphorous!Metabolism STAD 0.67 SLC34A3 0.003812225 FGF23!E>!SyndecanE2Emediated!signaling!events(RHOA,FGFR2,FGF19);FGF23!E>!SyndecanE3Emediated!signaling!events(FGFR2,FGF19,EGFR);SLC34A3!E>!Type!II!Na+/Pi!cotransporters(SLC34A3);FGF23!E>!FGF!signaling!pathway(CDH1,PIK3CA,FGF19,FGFR2)1 0.11223333 FGF23!E>!FGF19 1 !Combined!Heart!and!Skeletal!Defects STAD 1 0.002558855 CREBBP!E>!the!information!processing!pathway!at!the!ifn!beta!enhancer(IRF2,ARID1A);EP300!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CDH1,APC,AES,CDKN2A);EP300!E>!p73!transcription!factor!network(RNF43,CDK6,WWOX)0.650459398 0.04182609 CREBBP!E>!TP53,TP53,GATA40.26054167 !Specified!Hamartoses STAD 0.63 PTEN 1.38EE07 PTEN!E>!skeletal!muscle!hypertrophy!is!regulated!via!aktEmtor!pathway(PIK3CA,PTEN);PTEN!E>!regulation!of!eifE4e!and!p70s6!kinase(PIK3CA,PTEN);PTEN!E>!mtor!signaling!pathway(PTEN,PIK3CA);VHL!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53);PTEN!E>!RhoA!signaling!pathway(PTEN,RHOA,MAP2K4);PTEN!E>!pten!dependent!cell!cycle!arrest!and!apoptosis(PTEN,PIK3CA);PTEN!E>!Negative!regulation!of!the!PI3K/AKT!network(PTEN)0.719559223 0 PTEN!E>!PIK3CA;STK11!E>!EGFR1 !Li!Fraumeni!and!Related!Syndromes STAD 0.03 CDKN2A,TP53 1.68EE20 CDKN2A!E>!CEMYC!pathway(CDKN2A,FBXW7);TP53!E>!estrogen!responsive!protein!efp!controls!cell!cycle!and!breast!tumors!growth(TP53,CDK6);CDKN2A,TP53!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(CDKN2A,TP53);TP53!E>!telomeres!telomerase!cellular!aging!and!immortality(TP53,KRAS);CDKN2A!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CDH1,APC,AES,CDKN2A);TP53!E>!BARD1!signaling!events(CCNE1,TP53);TP53!E>!Transcriptional!!activation!of!!cell!cycle!inhibitor!p21(TP53);TP53,CHEK2!E>!PLK3!signaling!events(CCNE1,TP53);TP53!E>!p53!signaling!pathway(CCNE1,TP53);TP53!E>!p75(NTR)Emediated!signaling(PIK3CA,RHOA,TP53)0.693511719 0 CDKN2A!E>!PTEN;CHEK2!E>!CDK6,TP53,SMAD4;TP53!E>!PTEN1 !Chronic!Granulomatous!Disease UCEC 1 1 0.078663977 NCF2,NCF4,CYBA!E>!GMEB2(1.50eE02);NCF2,NCF4,CYBA!E>!ZNF263(1.79eE02);NCF2,CYBA!E>!IRAK1(2.62eE02);NCF2,NCF4,CYBB,CYBA!E>!HELZ2(1.44eE02);NCF2!E>!KRAS(4.41eE02);NCF2,NCF4,CYBB,CYBA!E>!ADAMDEC1(4.66eE02);NCF2,NCF4,CYBB,CYBA!E>!PTEN(1.31eE02);NCF2,NCF4,CYBB,CYBA!E>!PLAGL2(1.18eE02);NCF2,NCF4,CYBB,CYBA!E>!NFATC1(1.48eE02);NCF2,NCF4,CYBB,CYBA!E>!VDR(1.25eE02);NCF2,NCF4,CYBB,CYBA!E>!DNM2(1.64eE02);!E>!HAUS8(4.54eE02);NCF2,NCF4,CYBA!E>!ZGPAT(1.41eE02);NCF4,CYBB,CYBA!E>!NEK8(1.32eE02);NCF2,NCF4,CYBB,CYBA!E>!NFE2L2(1.19eE02);NCF2,NCF4,CYBB,CYBA!E>!CREBBP(1.30eE02);CYBA!E>!TMEM80(2.60eE02);NCF2,NCF4,CYBB,CYBA!E>!CTDP1(1.26eE02);NCF2,NCF4,CYBA!E>!CHMP2A(1.31eE02);CYBA!E>!RPLP2(4.66eE02);NCF2,NCF4,CYBB,CYBA!E>!SAP30BP(1.29eE02);CYBA!E>!POLD4(2.83eE02);NCF2,NCF4,CYBB,CYBA!E>!TPD52L2(1.15eE02);NCF4,CYBB,CYBA!E>!ADAM28(1.19eE02);NCF2,NCF4,CYBB,CYBA!E>!PQLC1(1.18eE02);NCF2,CYBB!E>!CTNNB1(3.35eE02);NCF2,NCF4,CYBB,CYBA!E>!SGK1(1.18eE02);NCF2,NCF4,CYBB,CYBA!E>!MYO9B(1.42eE02);NCF2,CYBB,CYBA!E>!NEU4(1.31eE02);NCF2,NCF4,CYBB,CYBA���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 1 !DiamondEBlackfan!Anemia UCEC 1 1.14EE06 RPS26,RPS24,RPS10,RPS17,RPS19,RPL5,RPL35A,RPS7,RPL11!E>!Regulation!of!gene!expression!in!beta!cells(RPL14,RPLP2,RPS5,FOXA2,RPL22);RPL11!E>!Validated!targets!of!CEMYC!transcriptional!activation(TAF4B,TERT,TP53,MYC,CREBBP)0.00122045 RPL5,RPS7!E>!NRAS(1.39eE03);RPS19,RPL35A,RPS7!E>!MYC(2.25eE04);RPS19!E>!IRAK1(3.08eE02);RPS26,RPS19,RPS10,RPL35A,RPS7!E>!TRIM28(1.64eE02);RPS26,RPS7!E>!HAUS8(1.64eE03);!E>!CCNE1(6.83eE03);RPS26,RPS7!E>!FTSJ2(1.78eE02);RPS26,RPS19,RPS10,RPL35A,RPS7!E>!NUDT1(2.14eE03);RPS26,RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7!E>!RPS5(3.05eE05);RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A!E>!RPLP2(3.96eE04);RPS19,RPS10!E>!POLD4(2.33eE02);RPS26,RPS19,RPS7!E>!TP53(1.72eE03);RPS26,RPS19,RPS7!E>!TACC3(8.31eE04);RPS26,RPS19,RPS7!E>!GEMIN4(2.74eE02);RPS26,RPS10,RPL35A!E>!RNMTL1(3.46eE04);RPS10,RPL35A!E>!ZNF497(1.11eE03);RPS26,RPS19,RPS10,RPL35A,RPS7!E>!TXNL4A(3.23eE02);RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7!E>!RPL22(3.92eE05);RPS10!E>!C9orf142(3.77eE03);!E>!TERT(1.23eE02);RPS19!E>!TRAF4(3.89eE02);RPS24,RPL5,RPS19,RPS10,RPL11,RPL35A,RPS7!E>!RPL14(2.97eE04);RPL11,RPL35A,RPS7!E>!RBMX(2.78eE02)0 RPL11!E>!FHIT;RPL35A!E>!MECOM;RPL5!E>!RPLP2;RPS10!E>!RPL14;RPS17!E>!RPS5;RPS19!E>!WWOX,FHIT;RPS24!E>!MECOM;RPS26!E>!RPLP2;RPS7!E>!RPL140.20041667 RPL11!E>!RPL14;RPL5!E>!TP53;RPS10!E>!ESR1;RPS17!E>!RPL22;RPS19!E>!RPS5;RPS24!E>!MYC,RPL14;RPS26!E>!ESR1;RPS7!E>!RPLP2!Inherited!Anomalies!of!the!Skin UCEC 1 TERT 3.29EE08 ATP2A2!E>!nfat!and!hypertrophy!of!the!heart!(NFATC1,CREBBP,PIK3R1,PIK3CA);TERT!E>!IL2!signaling!events!mediated!by!PI3K(PIK3CA,MYC,TERT,PIK3R1);TERT!E>!overview!of!telomerase!protein!component!gene!htert!transcriptional!regulation(MYC,TERT,TP53,MZF1,ESR1);TERT!E>!telomeres!telomerase!cellular!aging!and!immortality(MYC,TERT,TP53,RB1,KRAS);KRT1,TERT!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CTNNB1,CCND1,MYC,TERT);TERT!E>!role!of!nicotinic!acetylcholine!receptors!in!the!regulation!of!apoptosis(PIK3CA,TERT,PIK3R1);TERT!E>!Validated!targets!of!CEMYC!transcriptional!activation(TAF4B,TERT,TP53,MYC,CREBBP);TINF2,TERT,DKC1!E>!Regulation!of!Telomerase(CCND1,MYC,TERT,SIN3A,ESR1)0.183228108 1 1 !Combined!Heart!and!Skeletal!Defects UCEC 0.63 CREBBP 2.76EE30 CREBBP!E>!nfat!and!hypertrophy!of!the!heart!(NFATC1,CREBBP,PIK3R1,PIK3CA);EP300,CREBBP!E>!IFNEgamma!pathway(PIK3CA,CREBBP,PIK3R1);EP300,CREBBP!E>!Direct!p53!effectors(VDR,TP53,PIDD,RB1,PTEN,CREBBP);EP300,CREBBP!E>!FOXA1!transcription!factor!network(FOXA2,CREBBP,NKX3E1,ESR1);EP300,CREBBP!E>!transcription!regulation!by!methyltransferase!of!carm1(CREBBP,PRKAR1B);EP300!E>!cell!cycle:!g2/m!checkpoint(TP53,MYT1);EP300,CREBBP!E>!carm1!and!regulation!of!the!estrogen!receptor(CREBBP,ESR1);CREBBP!E>!wnt!signaling!pathway(CTNNB1,CCND1,MYC,CREBBP);EP300!E>!Validated!nuclear!estrogen!receptor!alpha!network(CCND1,MYC,UBE2M,ESR1);EP300!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CTNNB1,CCND1,MYC,TERT);EP300,CREBBP!E>!mechanism!of!gene!regulation!by!peroxisome!proliferators!via!ppara(MYC,PRKAR1B,RB1,CREBBP);CREBBP!E>!regulation!of!transcriptional!activity!by!pml(TP53,CREBBP,RB1);EP300,CREBBP!E>!E2F!transcription!factor!network(MYC,CCNE1,TRIM28,CREBBP,RB1);EP300,CREBBP!E>!ilE7!signal!transduc�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������1 0.33848148 1 !Hereditary!Sensory!Neuropathy UCEC 0.71 NEFL,DNM2 0.096035489 NTRK1!E>!Trk!receptor!signaling!mediated!by!PI3K!and!PLCEgamma(CCND1,PIK3CA,NRAS,PIK3R1,KRAS);NDRG1!E>!Direct!p53!effectors(VDR,TP53,PIDD,RB1,PTEN,CREBBP);NTRK1!E>!trka!receptor!signaling!pathway(PIK3CA,PIK3R1);NTRK1!E>!p73!transcription!factor!network(MYC,RNF43,RB1,WWOX);NTRK1!E>!p75(NTR)Emediated!signaling(PIK3CA,IRAK1,OMG,TP53,PIK3R1);DNM2!E>!PAR1Emediated!thrombin!signaling!events(PIK3CA,GNAQ,PIK3R1,DNM2);PMP22!E>!a6b1!and!a6b4!Integrin!signaling(ERBB2,PIK3CA,PIK3R1,ERBB3)0.130354569 1 1 !Li!Fraumeni!and!Related!Syndromes UCEC 0.63 TP53 1.96EE18 TP53!E>!chaperones!modulate!interferon!signaling!pathway(TP53,RB1);TP53!E>!Direct!p53!effectors(VDR,TP53,PIDD,RB1,PTEN,CREBBP);TP53!E>!LKB1!signaling!events(MYC,TP53,ESR1);TP53,CHEK2!E>!cell!cycle:!g2/m!checkpoint(TP53,MYT1);TP53!E>!estrogen!responsive!protein!efp!controls!cell!cycle!and!breast!tumors!growth(TP53,ESR1);TP53!E>!overview!of!telomerase!protein!component!gene!htert!transcriptional!regulation(MYC,TERT,TP53,MZF1,ESR1);CDKN2A!E>!Coregulation!of!Androgen!receptor!activity(CTNNB1,CCND1,CASP8,NKX3E1);CDKN2A,TP53!E>!Hypoxic!and!oxygen!homeostasis!regulation!of!HIFE1Ealpha(TP53,ARNT);TP53!E>!telomeres!telomerase!cellular!aging!and!immortality(MYC,TERT,TP53,RB1,KRAS);CDKN2A!E>!Regulation!of!nuclear!beta!catenin!signaling!and!target!gene!transcription(CTNNB1,CCND1,MYC,TERT);TP53!E>!btg!family!proteins!and!cell!cycle!regulation(CCND1,TP53,RB1);TP53!E>!Transcriptional!!activation!of!!cell!cycle!inhibitor!p21(TP53);TP53,CHEK2!E>!PLK3!signaling!events(CCNE1,TP53);TP53!E>!p53!signaling!pathway(CCND1,CCNE1,TP53,�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������0.458169296 0.57574242 1 !Lipoprotein!Deficiencies UCEC 1 1 0.064216795 MTTP,APOB,LCAT,SAR1B,APOA1!E>!A1BG(1.19eE02);MTTP,APOB,LCAT,SAR1B,APOA1!E>!SLC27A5(7.36eE03);MTTP,APOB,LCAT,SAR1B,APOA1!E>!HPD(1.19eE02);MTTP,APOB,LCAT,SAR1B,APOA1!E>!ATRN(2.26eE02);MTTP,APOB,LCAT,SAR1B,APOA1,ABCA1!E>!NEU4(1.39eE02)1 1 !

Page 134: Rachel D. Melamed

123

 Supplementary  Table  5:  ADAMS  results  for  comorbidity  with  acute  exacerbations  of  myasthenia  gravis.  Related  to  section  5.3.5.  

   

ICD$9&description

Have&disease&(case&set,&204&patients)

Incidence&(case&set)

Have&disease&(control&set,&2582&patients)

Incidence&(control&set)

Odds&ratio p$value

Carpal&tunnel&syndrome 9 0.044118 2 7.75E$04 56.96 2.47E$09Urinary&tract&infection&site&not&specified 26 0.127451 76 0.029435 4.33 5.95E$09Pneumonitis&due&to&inhalation&of&food&or&vomitus15 0.073529 21 0.008133 9.041 7.94E$09Anxiety&state&unspecified 14 0.068627 22 0.008521 8.054 7.39E$08Unspecified&essential&hypertension 53 0.259804 305 0.118125 2.199 9.73E$08Esophageal&reflux 19 0.093137 56 0.021689 4.294 8.60E$07Unspecified&pleural&effusion 12 0.058824 21 0.008133 7.232 1.55E$06Friedlander\'s&bacillus&infection&in&conditions&classified&elsewhere&and&of&unspecified&site6 0.029412 2 7.75E$04 37.97 3.55E$06Thyrotoxicosis&without&goiter&or&other&cause&and&without&thyrotoxic&crisis&or&storm6 0.029412 2 7.75E$04 37.97 3.55E$06Atrial&fibrillation 15 0.073529 41 0.015879 4.631 6.17E$06Other&specified&disorders&of&pancreatic&internal&secretion4 0.019608 0 0 $1 2.80E$05Other&specified&idiopathic&peripheral&neuropathy4 0.019608 0 0 $1 2.80E$05Pure&hypercholesterolemia 20 0.098039 83 0.032146 3.05 3.49E$05Hemorrhage&complicating&a&procedure 5 0.02451 2 7.75E$04 31.64 3.73E$05Long$term&(current)&use&of&steroids 5 0.02451 3 0.001162 21.09 9.37E$05Personal&history&of&noncompliance&with&medical&treatment&presenting&hazards&to&health7 0.034314 13 0.005035 6.815 3.47E$04Hematoma&complicating&a&procedure 4 0.019608 2 7.75E$04 25.31 3.73E$04Adrenal&cortical&steroids&causing&adverse&effects&in&therapeutic&use5 0.02451 5 0.001936 12.66 3.73E$04Nontoxic&uninodular&goiter 3 0.014706 0 0 $1 3.87E$04Chronic&lymphocytic&thyroiditis 3 0.014706 0 0 $1 3.87E$04Personal&history&of&malignant&neoplasm&of&bladder3 0.014706 0 0 $1 3.87E$04Personal&history&of&malignant&neoplasm&of&other&endocrine&glands&and&related&structures3 0.014706 0 0 $1 3.87E$04Embolism&and&thrombosis&of&other&specified&veins4 0.019608 3 0.001162 16.88 8.20E$04Depressive&disorder&not&elsewhere&classified 14 0.068627 62 0.024012 2.858 9.55E$04Toxic&diffuse&goiter&without&thyrotoxic&crisis&or&storm3 0.014706 1 3.87E$04 37.97 0.001465Unspecified&idiopathic&peripheral&neuropathy 3 0.014706 1 3.87E$04 37.97 0.001465Unspecified&disorder&of&optic&nerve&and&visual&pathways3 0.014706 1 3.87E$04 37.97 0.001465Personal&history&of&tobacco&use 8 0.039216 24 0.009295 4.219 0.001632Diabetes&mellitus&without&complication&type&i&not&stated&as&uncontrolled8 0.039216 25 0.009682 4.05 0.002022Hypertrophy&(benign)&of&prostate&without&urinary&obstruction5 0.02451 9 0.003486 7.032 0.002323Bipolar&disorder,&unspecified 4 0.019608 5 0.001936 10.13 0.002624Unspecified&disorder&of&thyroid 3 0.014706 2 7.75E$04 18.99 0.003465Retention&of&urine&unspecified 3 0.014706 2 7.75E$04 18.99 0.003465Other&specified&retention&of&urine 3 0.014706 2 7.75E$04 18.99 0.003465Migraine&unspecified&without&mention&of&intractable&migraine&without&mention&of&status&migrainosus4 0.019608 6 0.002324 8.438 0.004125Other&pulmonary&embolism&and&infarction 4 0.019608 6 0.002324 8.438 0.004125Unspecified&sleep&apnea 4 0.019608 6 0.002324 8.438 0.004125Anemia&unspecified 11 0.053922 53 0.020527 2.627 0.005828Methicillin&susceptible&staphylococcus&aureus 3 0.014706 3 0.001162 12.66 0.006558Obstructive&sleep&apnea&(adult)(pediatric) 3 0.014706 3 0.001162 12.66 0.006558Tracheostomy&status 4 0.019608 10 0.003873 5.063 0.015584

Page 135: Rachel D. Melamed

124

 8 REFERENCES

Adameyko,  Igor,  Francois  Lallemend,  Alessandro  Furlan,  Nikolay  Zinin,  Sergi  Aranda,  Satish  Srinivas  Kitambi,  Albert  Blanchart,  et  al.  2012.  “Sox2  and  Mitf  Cross-­‐Regulatory  Interactions  Consolidate  Progenitor  and  Melanocyte  Lineages  in  the  Cranial  Neural  Crest.”  Development  (Cambridge,  England)  139  (2):  397–410.  doi:10.1242/dev.065581.  http://dev.biologists.org/content/139/2/397.long.  

Akavia,  Uri  David,  Oren  Litvin,  Jessica  Kim,  Felix  Sanchez-­‐Garcia,  Dylan  Kotliar,  Helen  C  Causton,  Panisa  Pochanard,  Eyal  Mozes,  Levi  a  Garraway,  and  Dana  Pe’er.  2010.  “An  Integrated  Approach  to  Uncover  Drivers  of  Cancer.”  Cell  143  (6).  Elsevier  Inc.:  1005–17.  doi:10.1016/j.cell.2010.11.013.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3013278&tool=pmcentrez&rendertype=abstract.  

Akbani,  Rehan,  Patrick  Kwok  Shing  Ng,  Henrica  M  J  Werner,  Maria  Shahmoradgoli,  Fan  Zhang,  Zhenlin  Ju,  Wenbin  Liu,  et  al.  2014.  “A  Pan-­‐Cancer  Proteomic  Perspective  on  The  Cancer  Genome  Atlas.”  Nature  Communications  5  (May):  3887.  doi:10.1038/ncomms4887.  http://www.ncbi.nlm.nih.gov/pubmed/24871328.  

Anderson,  J  E.  1983.  “Seasonality  of  Symptomatic  Bacterial  Urinary  Infections  in  Women.”  Journal  of  Epidemiology  and  Community  Health  37  (4):  286–90.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1052926&tool=pmcentrez&rendertype=abstract.  

Armitage,  P,  and  R  Doll.  1954.  “The  Age  Distribution  of  Cancer  and  a  Multi-­‐Stage  Theory  of  Carcinogenesis.”  British  Journal  of  Cancer  VIII  (I):  1–12.  

Arnesen,  T,  D  Gromyko,  F  Pendino,  A  Ryningen,  J  E  Varhaug,  and  J  R  Lillehaug.  2006.  “Induction  of  Apoptosis  in  Human  Cells  by  RNAi-­‐Mediated  Knockdown  of  hARD1  and  NATH,  Components  of  the  Protein  N-­‐Alpha-­‐Acetyltransferase  Complex.”  Oncogene  25  (31):  4350–60.  doi:10.1038/sj.onc.1209469.  http://dx.doi.org/10.1038/sj.onc.1209469.  

Aydin,  Iraz  T.,  Rachel  D.  Melamed,  Sarah  J.  Adams,  Mireia  Castillo-­‐Martin,  Ahu  Demir,  Diana  Bryk,  Georg  Brunner,  et  al.  2014.  “FBXW7  Mutations  in  Melanoma  and  a  New  Therapeutic  Paradigm.”  Journal  of  the  National  Cancer  Institute  106  (6).  

Benjamini,  Y,  and  Y  Hochberg.  1995.  “Controlling  the  False  Discovery  Rate:  A  Practical  and  Powerful  Approach  to  Multiple  Testing.”  Journal  of  the  Royal  Statistical  Society.  Series  B  ….  http://www.jstor.org/stable/2346101.  

Beroukhim,  Rameen,  Gad  Getz,  Leia  Nghiemphu,  Jordi  Barretina,  Teli  Hsueh,  David  Linhart,  Igor  Vivanco,  et  al.  2007.  “Assessing  the  Significance  of  Chromosomal  Aberrations  in  Cancer:  Methodology  and  Application  to  Glioma.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  104  (50):  20007–12.  doi:10.1073/pnas.0710052104.  

Page 136: Rachel D. Melamed

125

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2148413&tool=pmcentrez&rendertype=abstract.  

Blair,  David  R.,  Christopher  S.  Lyttle,  Jonathan  M.  Mortensen,  Charles  F.  Bearden,  Anders  Boeck  Jensen,  Hossein  Khiabanian,  Rachel  Melamed,  et  al.  2013.  “A  Nondegenerate  Code  of  Deleterious  Variants  in  Mendelian  Loci  Contributes  to  Complex  Disease  Risk.”  Cell  155  (1).  Elsevier:  70–80.  doi:10.1016/j.cell.2013.08.030.  http://www.cell.com/fulltext/S0092-­‐8674(13)01024-­‐6.  

Bobak,  M.  2001.  “The  Seasonality  of  Live  Birth  Is  Strongly  Influenced  by  Socio-­‐Demographic  Factors.”  Human  Reproduction  16  (7):  1512–17.  doi:10.1093/humrep/16.7.1512.  http://humrep.oxfordjournals.org/content/16/7/1512.long.  

Boca,  Simina  M,  Kenneth  W  Kinzler,  Victor  E  Velculescu,  Bert  Vogelstein,  and  Giovanni  Parmigiani.  2010.  “Patient-­‐Oriented  Gene  Set  Analysis  for  Cancer  Mutation  Data.”  Genome  Biology  11  (11).  BioMed  Central  Ltd:  R112.  doi:10.1186/gb-­‐2010-­‐11-­‐11-­‐r112.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3156951&tool=pmcentrez&rendertype=abstract.  

Bourne,  T  David,  and  David  Schiff.  2010.  “Update  on  Molecular  Findings,  Management  and  Outcome  in  Low-­‐Grade  Gliomas.”  Nature  Reviews.  Neurology  6  (12).  Nature  Publishing  Group:  695–701.  doi:10.1038/nrneurol.2010.159.  http://www.ncbi.nlm.nih.gov/pubmed/21045797.  

Bowman,  A.  W.,  and  A.  Azzalini.  1997.  Applied  Smoothing  Techniques  for  Data  Analysis.  New  York:  Oxford  University  Press.  

Brauer,  Patrick  M,  and  Angela  L  Tyner.  2010.  “Building  a  Better  Understanding  of  the  Intracellular  Tyrosine  Kinase  PTK6  -­‐  BRK  by  BRK.”  Biochimica  et  Biophysica  Acta  1806  (1):  66–73.  doi:10.1016/j.bbcan.2010.02.003.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2885473&tool=pmcentrez&rendertype=abstract.  

Brennan,  Cameron  W,  Roel  G  W  Verhaak,  Aaron  McKenna,  Benito  Campos,  Houtan  Noushmehr,  Sofie  R  Salama,  Siyuan  Zheng,  et  al.  2013.  “The  Somatic  Genomic  Landscape  of  Glioblastoma.”  Cell  155  (2):  462–77.  doi:10.1016/j.cell.2013.09.034.  http://www.ncbi.nlm.nih.gov/pubmed/24120142.  

Brinkman,  Ryan  R,  Marie-­‐Pierre  Dubé,  Guy  A  Rouleau,  Andrew  C  Orr,  and  Mark  E  Samuels.  2006.  “Human  Monogenic  Disorders  -­‐  a  Source  of  Novel  Drug  Targets.”  Nature  Reviews.  Genetics  7  (4):  249–60.  doi:10.1038/nrg1828.  

Brown,  G,  C  Wallin,  T  Tatusova,  K  Pruitt,  and  D  Maglott.  2005.  “Gene  Help:  Integrated  Access  to  Genes  of  Genomes  in  the  Reference  Sequence  Collection.”  Bethesda,  MD:  National  Center  for  Biotechnology  Information.  

Bunt,  Jens,  Talitha  G  de  Haas,  Nancy  E  Hasselt,  Danny  A  Zwijnenburg,  Jan  Koster,  Rogier  Versteeg,  and  Marcel  Kool.  2010.  “Regulation  of  Cell  Cycle  Genes  and  Induction  of  

Page 137: Rachel D. Melamed

126

Senescence  by  Overexpression  of  OTX2  in  Medulloblastoma  Cell  Lines.”  Molecular  Cancer  Research :  MCR  8  (10):  1344–57.  doi:10.1158/1541-­‐7786.MCR-­‐09-­‐0546.  http://mcr.aacrjournals.org/content/8/10/1344.long.  

Burns,  Jane  C,  Daniel  R  Cayan,  Garrick  Tong,  Emelia  V  Bainto,  Christena  L  Turner,  Hiroko  Shike,  Tomisaku  Kawasaki,  Yosikazu  Nakamura,  Mayumi  Yashiro,  and  Hiroshi  Yanagawa.  2005.  “Seasonality  and  Temporal  Clustering  of  Kawasaki  Syndrome.”  Epidemiology  (Cambridge,  Mass.)  16  (2):  220–25.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2894624&tool=pmcentrez&rendertype=abstract.  

Cao,  Juxiang,  Lixin  Wan,  Elke  Hacker,  Xiangpeng  Dai,  Stefania  Lenna,  Celia  Jimenez-­‐Cervantes,  Yongjun  Wang,  et  al.  2013.  “MC1R  Is  a  Potent  Regulator  of  PTEN  after  UV  Exposure  in  Melanocytes.”  Molecular  Cell  51  (4):  409–22.  doi:10.1016/j.molcel.2013.08.010.  http://www.sciencedirect.com/science/article/pii/S1097276513005832.  

Cerami,  Ethan,  Emek  Demir,  Nikolaus  Schultz,  Barry  S  Taylor,  and  Chris  Sander.  2010.  “Automated  Network  Analysis  Identifies  Core  Pathways  in  Glioblastoma.”  PloS  One  5  (2):  e8918.  doi:10.1371/journal.pone.0008918.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2820542&tool=pmcentrez&rendertype=abstract.  

Challa-­‐Malladi,  Madhavi,  Yen  K  Lieu,  Olivia  Califano,  Antony  B  Holmes,  Govind  Bhagat,  Vundavalli  V  Murty,  David  Dominguez-­‐Sola,  Laura  Pasqualucci,  and  Riccardo  Dalla-­‐Favera.  2011.  “Combined  Genetic  Inactivation  of  β2-­‐Microglobulin  and  CD58  Reveals  Frequent  Escape  from  Immune  Recognition  in  Diffuse  Large  B  Cell  Lymphoma.”  Cancer  Cell  20  (6).  Elsevier:  728–40.  doi:10.1016/j.ccr.2011.11.006.  http://www.cell.com/article/S1535610811004375/fulltext.  

Ciriello,  Giovanni,  Ethan  G.  Cerami,  Chris  Sander,  and  Nikolaus  Schultz.  2011.  “Mutual  Exclusivity  Analysis  Identifies  Oncogenic  Network  Modules.”  Genome  Research  22  (2):  398–406.  doi:10.1101/gr.125567.111.  http://genome.cshlp.org/cgi/content/abstract/22/2/398.  

Ciriello,  Giovanni,  Martin  L  Miller,  Bülent  Arman  Aksoy,  Yasin  Senbabaoglu,  Nikolaus  Schultz,  and  Chris  Sander.  2013.  “Emerging  Landscape  of  Oncogenic  Signatures  across  Human  Cancers.”  Nature  Genetics  45  (10).  Nature  Publishing  Group,  a  division  of  Macmillan  Publishers  Limited.  All  Rights  Reserved.:  1127–33.  doi:10.1038/ng.2762.  http://dx.doi.org/10.1038/ng.2762.  

Clement,  Virginie,  Pilar  Sanchez,  Nicolas  de  Tribolet,  Ivan  Radovanovic,  and  Ariel  Ruiz  i  Altaba.  2007.  “HEDGEHOG-­‐GLI1  Signaling  Regulates  Human  Glioma  Growth,  Cancer  Stem  Cell  Self-­‐Renewal,  and  Tumorigenicity.”  Current  Biology :  CB  17  (2):  165–72.  doi:10.1016/j.cub.2006.11.033.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1855204&tool=pmcentrez&rendertype=abstract.  

Page 138: Rachel D. Melamed

127

Coloma,  Preciosa  M,  Vera  E  Valkhoff,  Giampiero  Mazzaglia,  Malene  Schou  Nielsson,  Lars  Pedersen,  Mariam  Molokhia,  Mees  Mosseveld,  et  al.  2013.  “Identification  of  Acute  Myocardial  Infarction  from  Electronic  Healthcare  Records  Using  Different  Disease  Coding  Systems:  A  Validation  Study  in  Three  European  Countries.”  BMJ  Open  3  (6).  doi:10.1136/bmjopen-­‐2013-­‐002862.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3686251&tool=pmcentrez&rendertype=abstract.  

“Comprehensive  Genomic  Characterization  Defines  Human  Glioblastoma  Genes  and  Core  Pathways.”  2008.  Nature  455  (7216):  1061–68.  doi:10.1038/nature07385.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2671642&tool=pmcentrez&rendertype=abstract.  

Consortium,  The  Fantom,  Riken  Pmi,  and  Clst  Dgt.  2014.  “A  Promoter-­‐Level  Mammalian  Expression  Atlas.”  Nature  507  (7493).  Nature  Publishing  Group:  462–70.  doi:10.1038/nature13182.  http://www.ncbi.nlm.nih.gov/pubmed/24670764.  

Cufi,  Perrine,  Nadine  Dragin,  Julia  Miriam  Weiss,  Pilar  Martinez-­‐Martinez,  Marc  H  De  Baets,  Régine  Roussin,  Elie  Fadel,  Sonia  Berrih-­‐Aknin,  and  Rozen  Le  Panse.  2013.  “Implication  of  Double-­‐Stranded  RNA  Signaling  in  the  Etiology  of  Autoimmune  Myasthenia  Gravis.”  Annals  of  Neurology  73  (2):  281–93.  doi:10.1002/ana.23791.  http://www.ncbi.nlm.nih.gov/pubmed/23280437.  

Dai,  Mu-­‐Shui,  and  Hua  Lu.  2004.  “Inhibition  of  MDM2-­‐Mediated  p53  Ubiquitination  and  Degradation  by  Ribosomal  Protein  L5.”  The  Journal  of  Biological  Chemistry  279  (43):  44475–82.  doi:10.1074/jbc.M403722200.  http://www.jbc.org/content/279/43/44475.  

Davies,  Helen,  Graham  R  Bignell,  Charles  Cox,  Philip  Stephens,  Sarah  Edkins,  Sheila  Clegg,  Jon  Teague,  et  al.  2002.  “Mutations  of  the  BRAF  Gene  in  Human  Cancer.”  Nature  417  (6892):  949–54.  doi:10.1038/nature00766.  http://www.ncbi.nlm.nih.gov/pubmed/12068308.  

DeBerardinis,  Ralph  J.,  Nabil  Sayed,  Dara  Ditsworth,  and  Craig  B.  Thompson.  2008.  “Brick  by  Brick:  Metabolism  and  Tumor  Cell  Growth.”  Current  Opinion  in  Genetics  and  Development  18:  54–61.  doi:10.1016/j.gde.2008.02.003.  

Demokan,  Semra,  Alice  Y  Chuang,  Kavita  M  Pattani,  David  Sidransky,  Wayne  Koch,  and  Joseph  A  Califano.  2014.  “Validation  of  Nucleolar  Protein  4  as  a  Novel  Methylated  Tumor  Suppressor  Gene  in  Head  and  Neck  Cancer.”  Oncology  Reports  31  (2):  1014–20.  doi:10.3892/or.2013.2927.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3896520&tool=pmcentrez&rendertype=abstract.  

Denny,  Joshua  C,  Lisa  Bastarache,  Marylyn  D  Ritchie,  Robert  J  Carroll,  Raquel  Zink,  Jonathan  D  Mosley,  Julie  R  Field,  et  al.  2013.  “Systematic  Comparison  of  Phenome-­‐Wide  Association  Study  of  Electronic  Medical  Record  Data  and  Genome-­‐Wide  Association  

Page 139: Rachel D. Melamed

128

Study  Data.”  Nature  Biotechnology  31  (12).  Nature  Publishing  Group:  1102–11.  doi:10.1038/nbt.2749.  http://www.ncbi.nlm.nih.gov/pubmed/24270849.  

Druker,  B  J,  M  Talpaz,  D  J  Resta,  B  Peng,  E  Buchdunger,  J  M  Ford,  N  B  Lydon,  et  al.  2001.  “Efficacy  and  Safety  of  a  Specific  Inhibitor  of  the  BCR-­‐ABL  Tyrosine  Kinase  in  Chronic  Myeloid  Leukemia.”  The  New  England  Journal  of  Medicine.  Vol.  344.  doi:10.1056/NEJM200104053441401.  

Eber,  Michael  R,  Michelle  Shardell,  Marin  L  Schweizer,  Ramanan  Laxminarayan,  and  Eli  N  Perencevich.  2011.  “Seasonal  and  Temperature-­‐Associated  Increases  in  Gram-­‐Negative  Bacterial  Bloodstream  Infections  among  Hospitalized  Patients.”  PloS  One  6  (9):  e25298.  doi:10.1371/journal.pone.0025298.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3180381&tool=pmcentrez&rendertype=abstract.  

Elder,  David  E.  2010.  “Dysplastic  Naevi:  An  Update.”  Histopathology.  

Falagas,  M  E,  G  Peppas,  D  K  Matthaiou,  D  E  Karageorgopoulos,  N  Karalis,  and  G  Theocharis.  2009.  “Effect  of  Meteorological  Variables  on  the  Incidence  of  Lower  Urinary  Tract  Infections.”  European  Journal  of  Clinical  Microbiology  &  Infectious  Diseases :  Official  Publication  of  the  European  Society  of  Clinical  Microbiology  28  (6):  709–12.  doi:10.1007/s10096-­‐008-­‐0679-­‐z.  http://www.ncbi.nlm.nih.gov/pubmed/19104854.  

Fine,  Jo  David,  Lorraine  B.  Johnson,  Madeline  Weiner,  Kuo  Ping  Li,  and  Chirayath  Suchindran.  2009.  “Epidermolysis  Bullosa  and  the  Risk  of  Life-­‐Threatening  Cancers:  The  National  EB  Registry  Experience,  1986-­‐2006.”  Journal  of  the  American  Academy  of  Dermatology  60  (2):  203–11.  

Finks,  JF,  NH  Osborne,  and  JD  Birkmeyer.  2011.  “Trends  in  Hospital  Volume  and  Operative  Mortality  for  High-­‐Risk  Surgery.”  New  England  Journal  of  …  364:  2128–37.  http://www.nejm.org/doi/full/10.1056/NEJMsa1010705#t=articleBackground.  

Frankel,  Mika,  Gabriel  Bekö,  Michael  Timm,  Sine  Gustavsen,  Erik  Wind  Hansen,  and  Anne  Mette  Madsen.  2012.  “Seasonal  Variations  of  Indoor  Microbial  Exposures  and  Their  Relation  to  Temperature,  Relative  Humidity,  and  Air  Exchange  Rate.”  Applied  and  Environmental  Microbiology  78  (23):  8289–97.  doi:10.1128/AEM.02069-­‐12.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3497365&tool=pmcentrez&rendertype=abstract.  

Friend,  Stephen  H,  R  Bernards,  S  Rogelj,  R  Weinberg,  JM  Rapaport,  DM  Albert,  and  TP  Dryja.  2014.  “A  Human  DNA  Segment  with  Properties  of  the  Gene  That  Predisposes  to  Retinoblastoma  and  Osteosarcoma.”  Accessed  June  25.  http://www.nature.com/scitable/content/A-­‐human-­‐DNA-­‐segment-­‐with-­‐properties-­‐of-­‐11477.  

Gallerani,  Massimo,  Benedetta  Boari,  Fabio  Manfredini,  and  Roberto  Manfredini.  2011.  “Seasonal  Variation  in  Heart  Failure  Hospitalization.”  Clinical  Cardiology  34  (6):  389–94.  doi:10.1002/clc.20895.  http://www.ncbi.nlm.nih.gov/pubmed/21538387.  

Page 140: Rachel D. Melamed

129

Goebel,  B.,  Z.  Dawy,  J.  Hagenauer,  and  J.C.  Mueller.  “An  Approximation  to  the  Distribution  of  Finite  Sample  Size  Mutual  Information  Estimates.”  IEEE  International  Conference  on  Communications,  2005.  ICC  2005.  2005  2  (4).  Ieee:  1102–6.  doi:10.1109/ICC.2005.1494518.  http://ieeexplore.ieee.org/lpdocs/epic03/wrapper.htm?arnumber=1494518.  

Gottesman,  Omri,  Helena  Kuivaniemi,  Gerard  Tromp,  W  Andrew  Faucett,  Rongling  Li,  Teri  A  Manolio,  Saskia  C  Sanderson,  et  al.  2013.  “The  Electronic  Medical  Records  and  Genomics  (eMERGE)  Network:  Past,  Present,  and  Future.”  Genetics  in  Medicine :  Official  Journal  of  the  American  College  of  Medical  Genetics  15  (10).  American  College  of  Medical  Genetics  and  Genomics:  761–71.  doi:10.1038/gim.2013.72.  http://dx.doi.org/10.1038/gim.2013.72.  

Gudbjartsson,  Daniel  F,  Patrick  Sulem,  Simon  N  Stacey,  Alisa  M  Goldstein,  Thorunn  Rafnar,  Bardur  Sigurgeirsson,  Kristrun  R  Benediktsdottir,  et  al.  2008.  “ASIP  and  TYR  Pigmentation  Variants  Associate  with  Cutaneous  Melanoma  and  Basal  Cell  Carcinoma.”  Nature  Genetics  40  (7).  Nature  Publishing  Group:  886–91.  doi:10.1038/ng.161.  http://dx.doi.org/10.1038/ng.161.  

Hanahan,  Douglas,  and  Robert  A  Weinberg.  2011.  “Review  Hallmarks  of  Cancer :  The  Next  Generation.”  Cell  144  (5).  Elsevier  Inc.:  646–74.  doi:10.1016/j.cell.2011.02.013.  http://dx.doi.org/10.1016/j.cell.2011.02.013.  

Hayashi,  Reiko,  Yuya  Goto,  Ryuji  Ikeda,  Kazunari  K  Yokoyama,  and  Kenichi  Yoshida.  2006.  “CDCA4  Is  an  E2F  Transcription  Factor  Family-­‐Induced  Nuclear  Factor  That  Regulates  E2F-­‐Dependent  Transcriptional  Activation  and  Cell  Proliferation.”  The  Journal  of  Biological  Chemistry  281  (47):  35633–48.  doi:10.1074/jbc.M603800200.  http://www.ncbi.nlm.nih.gov/pubmed/16984923.  

Hoadley,  Katherine  a.,  Christina  Yau,  Denise  M.  Wolf,  Andrew  D.  Cherniack,  David  Tamborero,  Sam  Ng,  Max  D.M.  Leiserson,  et  al.  2014.  “Multiplatform  Analysis  of  12  Cancer  Types  Reveals  Molecular  Classification  within  and  across  Tissues  of  Origin.”  Cell  158  (4).  Elsevier  Inc.:  929–44.  doi:10.1016/j.cell.2014.06.049.  http://linkinghub.elsevier.com/retrieve/pii/S0092867414008769.  

Hoehndorf,  Robert,  Paul  N  Schofield,  and  Georgios  V  Gkoutos.  2013.  “An  Integrative,  Translational  Approach  to  Understanding  Rare  and  Orphan  Genetically  Based  Diseases.”  Interface  Focus  3  (2):  20120055.  doi:10.1098/rsfs.2012.0055.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3638468&tool=pmcentrez&rendertype=abstract.  

Hoek,  Keith  S,  Natalie  C  Schlegel,  Ossia  M  Eichhoff,  Daniel  S  Widmer,  Christian  Praetorius,  Steingrimur  O  Einarsson,  Sigridur  Valgeirsdottir,  et  al.  2008.  “Novel  MITF  Targets  Identified  Using  a  Two-­‐Step  DNA  Microarray  Strategy.”  Pigment  Cell  &  Melanoma  Research  21  (6):  665–76.  doi:10.1111/j.1755-­‐148X.2008.00505.x.  http://www.ncbi.nlm.nih.gov/pubmed/19067971.  

Page 141: Rachel D. Melamed

130

Hofree,  Matan,  John  P  Shen,  Hannah  Carter,  Andrew  Gross,  and  Trey  Ideker.  2013.  “Network-­‐Based  Stratification  of  Tumor  Mutations.”  Nature  Methods  10  (11):  1108–15.  doi:10.1038/nmeth.2651.  http://www.ncbi.nlm.nih.gov/pubmed/24037242.  

Holmes,  Antony  B,  Alexander  Hawson,  Feng  Liu,  Carol  Friedman,  Hossein  Khiabanian,  and  Raul  Rabadan.  2011.  “Discovering  Disease  Associations  by  Integrating  Electronic  Clinical  Data  and  Medical  Literature.”  PloS  One  6  (6):  e21132.  doi:10.1371/journal.pone.0021132.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3121722&tool=pmcentrez&rendertype=abstract.  

Hwang,  Harry  C,  and  Bruce  E  Clurman.  2005.  “Cyclin  E  in  Normal  and  Neoplastic  Cell  Cycles.”  Oncogene  24  (17):  2776–86.  doi:10.1038/sj.onc.1208613.  http://dx.doi.org/10.1038/sj.onc.1208613.  

Kamburov,  Atanas,  Ulrich  Stelzl,  Hans  Lehrach,  and  Ralf  Herwig.  2013.  “The  ConsensusPathDB  Interaction  Database:  2013  Update.”  Nucleic  Acids  Research  41  (Database  issue):  D793–800.  doi:10.1093/nar/gks1055.  http://nar.oxfordjournals.org/content/41/D1/D793.  

Khiabanian,  Hossein,  Antony  B  Holmes,  Brendan  J  Kelly,  Mrinalini  Gururaj,  George  Hripcsak,  and  Raul  Rabadan.  2010.  “Signs  of  the  2009  Influenza  Pandemic  in  the  New  York-­‐Presbyterian  Hospital  Electronic  Health  Records.”  Edited  by  Nancy  Mock.  PloS  One  5  (9).  Public  Library  of  Science:  8.  doi:10.1371/journal.pone.0012658.  http://dx.plos.org/10.1371/journal.pone.0012658.  

Kishimoto,  Masahiro,  Takashi  Kohno,  Koji  Okudela,  Ayaka  Otsuka,  Hiroki  Sasaki,  Chikako  Tanabe,  Tokuki  Sakiyama,  et  al.  2005.  “Mutations  and  Deletions  of  the  CBP  Gene  in  Human  Lung  Cancer.”  Clinical  Cancer  Research :  An  Official  Journal  of  the  American  Association  for  Cancer  Research  11  (2  Pt  1):  512–19.  http://www.ncbi.nlm.nih.gov/pubmed/15701835.  

Knudson,  AG.  2001.  “Two  Genetic  Hits  (more  or  Less)  to  Cancer.”  Nature  Reviews  Cancer  1  (November):  637–41.  http://www.nature.com/nrc/journal/v1/n2/abs/nrc1101-­‐157a.html.  

Kobayashi,  Koichi  S,  and  Peter  J  van  den  Elsen.  2012.  “NLRC5:  A  Key  Regulator  of  MHC  Class  I-­‐Dependent  Immune  Responses.”  Nature  Reviews.  Immunology  12  (12).  Nature  Publishing  Group,  a  division  of  Macmillan  Publishers  Limited.  All  Rights  Reserved.:  813–20.  doi:10.1038/nri3339.  http://dx.doi.org/10.1038/nri3339.  

Kullback,  S.  2012.  “Information  Theory  and  Statistics.”  Accessed  October  21.  http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.14.5452.  

Lambe,  M,  P  Blomqvist,  and  R  Bellocco.  2003.  “Seasonal  Variation  in  the  Diagnosis  of  Cancer:  A  Study  Based  on  National  Cancer  Registration  in  Sweden.”  British  Journal  of  Cancer  88  (9).  Cancer  Research  UK:  1358–60.  doi:10.1038/sj.bjc.6600901.  http://dx.doi.org/10.1038/sj.bjc.6600901.  

Page 142: Rachel D. Melamed

131

Lambert,  G  W,  C  Reid,  D  M  Kaye,  G  L  Jennings,  and  M  D  Esler.  2002.  “Effect  of  Sunlight  and  Season  on  Serotonin  Turnover  in  the  Brain.”  Lancet  360  (9348):  1840–42.  http://www.ncbi.nlm.nih.gov/pubmed/12480364.  

Law,  Charity  W,  Yunshun  Chen,  Wei  Shi,  and  Gordon  K  Smyth.  2014.  “Voom:  Precision  Weights  Unlock  Linear  Model  Analysis  Tools  for  RNA-­‐Seq  Read  Counts.”  Genome  Biology  15  (2):  R29.  doi:10.1186/gb-­‐2014-­‐15-­‐2-­‐r29.  http://genomebiology.com/2014/15/2/R29.  

Lawrence,  Michael  S,  Petar  Stojanov,  Craig  H  Mermel,  James  T  Robinson,  Levi  A  Garraway,  Todd  R  Golub,  Matthew  Meyerson,  Stacey  B  Gabriel,  Eric  S  Lander,  and  Gad  Getz.  2014.  “Discovery  and  Saturation  Analysis  of  Cancer  Genes  across  21  Tumour  Types.”  Nature  505  (7484).  Nature  Publishing  Group,  a  division  of  Macmillan  Publishers  Limited.  All  Rights  Reserved.:  495–501.  doi:10.1038/nature12912.  http://dx.doi.org/10.1038/nature12912.  

Lawrence,  Michael  S,  Petar  Stojanov,  Paz  Polak,  Gregory  V  Kryukov,  Kristian  Cibulskis,  Andrey  Sivachenko,  Scott  L  Carter,  et  al.  2013.  “Mutational  Heterogeneity  in  Cancer  and  the  Search  for  New  Cancer-­‐Associated  Genes.”  Nature  499  (7457):  214–18.  doi:10.1038/nature12213.  http://www.ncbi.nlm.nih.gov/pubmed/23770567.  

Lee,  D-­‐S,  J  Park,  K  a  Kay,  N  a  Christakis,  Z  N  Oltvai,  and  A-­‐L  Barabási.  2008.  “The  Implications  of  Human  Metabolic  Network  Topology  for  Disease  Comorbidity.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  105  (29):  9880–85.  doi:10.1073/pnas.0802208105.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2481357&tool=pmcentrez&rendertype=abstract.  

Lee,  Insuk,  U  Martin  Blom,  Peggy  I  Wang,  Jung  Eun  Shim,  and  Edward  M  Marcotte.  2011.  “Prioritizing  Candidate  Disease  Genes  by  Network-­‐Based  Boosting  of  Genome-­‐Wide  Association  Data.”  Genome  Research  21  (7):  1109–21.  doi:10.1101/gr.118992.110.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3129253&tool=pmcentrez&rendertype=abstract.  

Leiserson,  Mark  D  M,  Fabio  Vandin,  Hsin-­‐Ta  Wu,  Jason  R  Dobson,  Jonathan  V  Eldridge,  Jacob  L  Thomas,  Alexandra  Papoutsaki,  et  al.  2014.  “Pan-­‐Cancer  Network  Analysis  Identifies  Combinations  of  Rare  Somatic  Mutations  across  Pathways  and  Protein  Complexes.”  Nature  Genetics,  December.  Nature  Publishing  Group.  doi:10.1038/ng.3168.  http://www.nature.com/doifinder/10.1038/ng.3168.  

Leiserson,  Mark  D.  M.,  Dima  Blokh,  Roded  Sharan,  and  Benjamin  J.  Raphael.  2013.  “Simultaneous  Identification  of  Multiple  Driver  Pathways  in  Cancer.”  Edited  by  Niko  Beerenwinkel.  PLoS  Computational  Biology  9  (5):  e1003054.  doi:10.1371/journal.pcbi.1003054.  http://dx.plos.org/10.1371/journal.pcbi.1003054.  

Levine,  Pamela  J,  Miriam  R  Elman,  Ravina  Kullar,  John  M  Townes,  David  T  Bearden,  Rowena  Vilches-­‐Tran,  Ian  McClellan,  and  Jessina  C  McGregor.  2013.  “Use  of  Electronic  Health  Record  Data  to  Identify  Skin  and  Soft  Tissue  Infections  in  Primary  Care  Settings:  A  

Page 143: Rachel D. Melamed

132

Validation  Study.”  BMC  Infectious  Diseases  13  (January):  171.  doi:10.1186/1471-­‐2334-­‐13-­‐171.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3637223&tool=pmcentrez&rendertype=abstract.  

Levy,  Carmit,  Mehdi  Khaled,  and  D.E.  David  E  Fisher.  2006.  “MITF:  Master  Regulator  of  Melanocyte  Development  and  Melanoma  Oncogene.”  Trends  in  Molecular  Medicine  12  (9).  Elsevier:  406–14.  doi:10.1016/j.molmed.2006.07.008.  http://linkinghub.elsevier.com/retrieve/pii/S1471491406001699.  

Li,  Bo,  and  Colin  N  Dewey.  2011.  “RSEM:  Accurate  Transcript  Quantification  from  RNA-­‐Seq  Data  with  or  without  a  Reference  Genome.”  BMC  Bioinformatics  12  (January):  323.  doi:10.1186/1471-­‐2105-­‐12-­‐323.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3163565&tool=pmcentrez&rendertype=abstract.  

Li,  Caiyun  G,  and  Michael  R  Eccles.  2012.  “PAX  Genes  in  Cancer;  Friends  or  Foes?”  Frontiers  in  Genetics  3  (January).  Frontiers:  6.  doi:10.3389/fgene.2012.00006.  http://journal.frontiersin.org/Journal/10.3389/fgene.2012.00006/abstract.  

Li,  Heng,  and  Richard  Durbin.  2009.  “Fast  and  Accurate  Short  Read  Alignment  with  Burrows-­‐Wheeler  Transform.”  Bioinformatics  (Oxford,  England)  25  (14):  1754–60.  doi:10.1093/bioinformatics/btp324.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2705234&tool=pmcentrez&rendertype=abstract.  

Li,  Ruwei,  Bo  Wan,  Jun  Zhou,  Yingli  Wang,  Ting  Luo,  Xiuting  Gu,  Fang  Chen,  and  Long  Yu.  2012.  “APC/C(Cdh1)  Targets  Brain-­‐Specific  Kinase  2  (BRSK2)  for  Degradation  via  the  Ubiquitin-­‐Proteasome  Pathway.”  Edited  by  Deanna  M.  Koepp.  PloS  One  7  (9).  Public  Library  of  Science:  e45932.  doi:10.1371/journal.pone.0045932.  http://dx.plos.org/10.1371/journal.pone.0045932.  

Llaverias,  Gemma,  Christiane  Danilo,  Isabelle  Mercier,  Kristin  Daumer,  Franco  Capozza,  Terence  M.  Williams,  Federica  Sotgia,  Michael  P.  Lisanti,  and  Philippe  G.  Frank.  2011.  “Role  of  Cholesterol  in  the  Development  and  Progression  of  Breast  Cancer.”  American  Journal  of  Pathology  178  (1):  402–12.  

Lomb,  N.  R.  1976.  “Least-­‐Squares  Frequency  Analysis  of  Unequally  Spaced  Data.”  Astrophysics  and  Space  Science  39  (2):  447–62.  doi:10.1007/BF00648343.  http://link.springer.com/10.1007/BF00648343.  

Love,  Michael  I,  Wolfgang  Huber,  and  Simon  Anders.  2014.  “Moderated  Estimation  of  Fold  Change  and  Dispersion  for  RNA-­‐Seq  Data  with  DESeq2.”  Genome  Biology  15  (12):  550.  doi:10.1186/s13059-­‐014-­‐0550-­‐8.  http://genomebiology.com/2014/15/12/550.  

Malkin,  D,  F  P  Li,  L  C  Strong,  J  F  Fraumeni,  C  E  Nelson,  D  H  Kim,  J  Kassel,  M  A  Gryka,  F  Z  Bischoff,  and  M  A  Tainsky.  1990.  “Germ  Line  p53  Mutations  in  a  Familial  Syndrome  of  

Page 144: Rachel D. Melamed

133

Breast  Cancer,  Sarcomas,  and  Other  Neoplasms.”  Science  (New  York,  N.Y.)  250  (4985):  1233–38.  http://www.ncbi.nlm.nih.gov/pubmed/1978757.  

Manfredi,  James  J.  2010.  “The  Mdm2-­‐p53  Relationship  Evolves:  Mdm2  Swings  Both  Ways  as  an  Oncogene  and  a  Tumor  Suppressor.”  Genes  &  Development  24  (15):  1580–89.  doi:10.1101/gad.1941710.  http://genesdev.cshlp.org/content/24/15/1580.long.  

Margolin,  Adam  A,  Ilya  Nemenman,  Katia  Basso,  Chris  Wiggins,  Gustavo  Stolovitzky,  Riccardo  Dalla  Favera,  and  Andrea  Califano.  2006.  “ARACNE:  An  Algorithm  for  the  Reconstruction  of  Gene  Regulatory  Networks  in  a  Mammalian  Cellular  Context.”  BMC  Bioinformatics  7  Suppl  1  (Suppl  1):  S7.  doi:10.1186/1471-­‐2105-­‐7-­‐S1-­‐S7.  http://www.ncbi.nlm.nih.gov/pubmed/16723010.  

Maslov,  Sergei,  and  Kim  Sneppen.  2002.  “Specificity  and  Stability  in  Topology  of  Protein  Networks.”  Science  296  (5569):  910–13.  doi:10.1126/science.1065103.  http://www.ncbi.nlm.nih.gov/pubmed/11988575.  

McCarty,  Catherine  A,  Rex  L  Chisholm,  Christopher  G  Chute,  Iftikhar  J  Kullo,  Gail  P  Jarvik,  Eric  B  Larson,  Rongling  Li,  et  al.  2011.  “The  eMERGE  Network:  A  Consortium  of  Biorepositories  Linked  to  Electronic  Medical  Records  Data  for  Conducting  Genomic  Studies.”  BMC  Medical  Genomics.  

McKusick-­‐Nathans  Institute  of  Genetic  Medicine,  Johns  Hopkins  University  (Baltimore,  MD).  “Online  Mendelian  Inheritance  in  Man,  OMIM®.”  Http://omim.org/.  

Melamed,  Rachel  D,  Hossein  Khiabanian,  and  Raul  Rabadan.  2014.  “Data-­‐Driven  Discovery  of  Seasonally  Linked  Diseases  from  an  Electronic  Health  Records  System.”  BMC  Bioinformatics  15  Suppl  6  (Suppl  6).  BioMed  Central  Ltd:  S3.  doi:10.1186/1471-­‐2105-­‐15-­‐S6-­‐S3.  http://www.ncbi.nlm.nih.gov/pubmed/25078762.  

Mermel,  Craig  H,  Steven  E  Schumacher,  Barbara  Hill,  Matthew  L  Meyerson,  Rameen  Beroukhim,  and  Gad  Getz.  2011.  “GISTIC2.0  Facilitates  Sensitive  and  Confident  Localization  of  the  Targets  of  Focal  Somatic  Copy-­‐Number  Alteration  in  Human  Cancers.”  Genome  Biology  12  (4):  R41.  doi:10.1186/gb-­‐2011-­‐12-­‐4-­‐r41.  http://genomebiology.com/2011/12/4/R41.  

Metral,  Sylvain,  Beata  Machnicka,  Sylvain  Bigot,  Yves  Colin,  Didier  Dhermy,  and  Marie-­‐Christine  Lecomte.  2009.  “AlphaII-­‐Spectrin  Is  Critical  for  Cell  Adhesion  and  Cell  Cycle.”  The  Journal  of  Biological  Chemistry  284  (4):  2409–18.  doi:10.1074/jbc.M801324200.  

Michaloglou,  Chrysiis,  Liesbeth  C  W  Vredeveld,  Maria  S  Soengas,  Christophe  Denoyelle,  Thomas  Kuilman,  Chantal  M  A  M  van  der  Horst,  Donné  M  Majoor,  Jerry  W  Shay,  Wolter  J  Mooi,  and  Daniel  S  Peeper.  2005.  “BRAFE600-­‐Associated  Senescence-­‐like  Cell  Cycle  Arrest  of  Human  Naevi.”  Nature  436  (7051):  720–24.  doi:10.1038/nature03890.  http://www.ncbi.nlm.nih.gov/pubmed/16079850.  

Miller,  Christopher  A,  Stephen  H  Settle,  Erik  P  Sulman,  Kenneth  D  Aldape,  and  Aleksandar  Milosavljevic.  2011.  “Discovering  Functional  Modules  by  Identifying  Recurrent  and  

Page 145: Rachel D. Melamed

134

Mutually  Exclusive  Mutational  Patterns  in  Tumors.”  BMC  Medical  Genomics  4  (1):  34.  doi:10.1186/1755-­‐8794-­‐4-­‐34.  http://www.biomedcentral.com/1755-­‐8794/4/34.  

Miller,  R  W,  and  J  H  Rubinstein.  1995.  “Tumors  in  Rubinstein-­‐Taybi  Syndrome.”  American  Journal  of  Medical  Genetics  56  (1):  112–15.  doi:10.1002/ajmg.1320560125.  http://www.ncbi.nlm.nih.gov/pubmed/7747773.  

Mo,  Qianxing,  Sijian  Wang,  Venkatraman  E  Seshan,  Adam  B  Olshen,  Nikolaus  Schultz,  Chris  Sander,  R  Scott  Powers,  Marc  Ladanyi,  and  Ronglai  Shen.  2013.  “Pattern  Discovery  and  Cancer  Gene  Identification  in  Integrated  Cancer  Genomic  Data.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  110  (11):  4245–50.  doi:10.1073/pnas.1208949110.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3600490&tool=pmcentrez&rendertype=abstract.  

Molendijk,  Marc  L,  Judith  P  M  Haffmans,  Boudewijn  a  a  Bus,  Philip  Spinhoven,  Brenda  W  J  H  Penninx,  Jos  Prickaerts,  Richard  C  Oude  Voshaar,  and  Bernet  M  Elzinga.  2012.  “Serum  BDNF  Concentrations  Show  Strong  Seasonal  Variation  and  Correlations  with  the  Amount  of  Ambient  Sunlight.”  PloS  One  7  (11):  e48046.  doi:10.1371/journal.pone.0048046.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3487856&tool=pmcentrez&rendertype=abstract.  

Mullighan,  Charles  G,  Jinghui  Zhang,  Lawryn  H  Kasper,  Stephanie  Lerach,  Debbie  Payne-­‐Turner,  Letha  A  Phillips,  Sue  L  Heatley,  et  al.  2011.  “CREBBP  Mutations  in  Relapsed  Acute  Lymphoblastic  Leukaemia.”  Nature  471  (7337).  Nature  Publishing  Group,  a  division  of  Macmillan  Publishers  Limited.  All  Rights  Reserved.:  235–39.  doi:10.1038/nature09727.  http://dx.doi.org/10.1038/nature09727.  

Noushmehr,  Houtan,  Daniel  J  Weisenberger,  Kristin  Diefes,  Heidi  S  Phillips,  Kanan  Pujara,  Benjamin  P  Berman,  Fei  Pan,  et  al.  2010.  “Identification  of  a  CpG  Island  Methylator  Phenotype  That  Defines  a  Distinct  Subgroup  of  Glioma.”  Cancer  Cell  17  (5).  Elsevier  Ltd:  510–22.  doi:10.1016/j.ccr.2010.03.017.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2872684&tool=pmcentrez&rendertype=abstract.  

Nowell,  P,  and  D  Hungerford.  1960.  “A  Minute  Chromosome  in  Human  Chronic  Granulocytic  Leukemia.”  Science  132:  1497.  

Oberg,  C,  J  Li,  A  Pauley,  E  Wolf,  M  Gurney,  and  U  Lendahl.  2001.  “The  Notch  Intracellular  Domain  Is  Ubiquitinated  and  Negatively  Regulated  by  the  Mammalian  Sel-­‐10  Homolog.”  The  Journal  of  Biological  Chemistry  276  (38):  35847–53.  doi:10.1074/jbc.M103992200.  

Onyango,  Patrick,  and  Andrew  P  Feinberg.  2011.  “A  Nucleolar  Protein,  H19  Opposite  Tumor  Suppressor  (HOTS),  Is  a  Tumor  Growth  Inhibitor  Encoded  by  a  Human  Imprinted  H19  Antisense  Transcript.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  108  (40):  16759–64.  doi:10.1073/pnas.1110904108.  

Page 146: Rachel D. Melamed

135

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3189046&tool=pmcentrez&rendertype=abstract.  

Palmieri,  Giuseppe,  Mariaelena  Capone,  Maria  Ascierto,  Giusy  Gentilcore,  David  F  Stroncek,  Milena  Casula,  Maria  Sini,  Marco  Palla,  Nicola  Mozzillo,  and  Paolo  a  Ascierto.  2009.  “Main  Roads  to  Melanoma.”  Journal  of  Translational  Medicine  7  (1):  86.  doi:10.1186/1479-­‐5876-­‐7-­‐86.  http://www.translational-­‐medicine.com/content/7/1/86.  

Park,  Juyong,  Deok-­‐Sun  Lee,  Nicholas  a  Christakis,  and  Albert-­‐László  Barabási.  2009.  “The  Impact  of  Cellular  Networks  on  Disease  Comorbidity.”  Molecular  Systems  Biology  5  (262):  262.  doi:10.1038/msb.2009.16.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2683720&tool=pmcentrez&rendertype=abstract.  

Pasqualucci,  Laura,  David  Dominguez-­‐Sola,  Annalisa  Chiarenza,  Giulia  Fabbri,  Adina  Grunn,  Vladimir  Trifonov,  Lawryn  H  Kasper,  et  al.  2011.  “Inactivating  Mutations  of  Acetyltransferase  Genes  in  B-­‐Cell  Lymphoma.”  Nature  471  (7337).  Nature  Publishing  Group:  189–95.  doi:10.1038/nature09730.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3271441&tool=pmcentrez&rendertype=abstract.  

Pei,  Lei,  Pingfang  Xie,  Enxiang  Zhou,  Qian  Yang,  Yi  Luo,  and  Zhonghua  Tang.  “Overexpression  of  DEP  Domain  Containing  mTOR-­‐Interacting  Protein  Correlates  with  Poor  Prognosis  in  Differentiated  Thyroid  Carcinoma.”  Molecular  Medicine  Reports  4  (5):  817–23.  doi:10.3892/mmr.2011.503.  http://www.ncbi.nlm.nih.gov/pubmed/21643629.  

Perencevich,  Eli  N.,  Jessina  C.  McGregor,  Michelle  Shardell,  Jon  P.  Furuno,  Anthony  D.  Harris,  J.  Glenn  Morris,  Jr.,  David  N.  Fisman,  and  Judith  A.  Johnson.  2008.  “Summer  Peaks  in  the  Incidences  of  Gram-­‐Negative  Bacterial  Infection  Among  Hospitalized  Patients  •.”  Infection  Control  and  Hospital  Epidemiology  29  (12):  1124–31.  doi:10.1086/592698.  http://www.jstor.org/stable/10.1086/592698.  

Poulikakos,  Poulikos  I.,  and  Neal  Rosen.  2011.  “Mutant  BRAF  Melanomas—Dependence  and  Resistance.”  Cancer  Cell  19  (1).  Elsevier  Inc.:  11–15.  doi:10.1016/j.ccr.2011.01.008.  http://www.ncbi.nlm.nih.gov/pubmed/21251612.  

Press,  WH.  1992.  “Numerical  Recipes  in  C :  The  Art  of  Scientific  Computing.”  In  ,  2nd  ed.  Cambridge ;  New  York:  Cambridge  University  Press.  

Querol,  Luis,  and  Isabel  Illa.  2013.  “Myasthenia  Gravis  and  the  Neuromuscular  Junction.”  Current  Opinion  in  Neurology  26  (5):  459–65.  doi:10.1097/WCO.0b013e328364c079.  http://www.ncbi.nlm.nih.gov/pubmed/23945282.  

Raimondi,  Sara,  Francesco  Sera,  Sara  Gandini,  Simona  Iodice,  Saverio  Caini,  Patrick  Maisonneuve,  and  Maria  Concetta  Fargnoli.  2008.  “MC1R  Variants,  Melanoma  and  Red  Hair  Color  Phenotype:  A  Meta-­‐Analysis.”  International  Journal  of  Cancer.  Journal  

Page 147: Rachel D. Melamed

136

International  Du  Cancer  122  (12):  2753–60.  doi:10.1002/ijc.23396.  http://www.ncbi.nlm.nih.gov/pubmed/18366057.  

Remke,  Marc,  Vijay  Ramaswamy,  John  Peacock,  David  J  H  Shih,  Christian  Koelsche,  Paul  a  Northcott,  Nadia  Hill,  et  al.  2013.  “TERT  Promoter  Mutations  Are  Highly  Recurrent  in  SHH  Subgroup  Medulloblastoma.”  Acta  Neuropathologica  126  (6):  917–29.  doi:10.1007/s00401-­‐013-­‐1198-­‐2.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3830749&tool=pmcentrez&rendertype=abstract.  

Rodó,  Xavier,  Joan  Ballester,  Dan  Cayan,  Marian  E  Melish,  Yoshikazu  Nakamura,  Ritei  Uehara,  and  Jane  C  Burns.  2011.  “Association  of  Kawasaki  Disease  with  Tropospheric  Wind  Patterns.”  Scientific  Reports  1  (January).  Nature  Publishing  Group:  152.  doi:10.1038/srep00152.  http://www.nature.com/srep/2011/111110/srep00152/full/srep00152.html.  

Sato,  S,  K  Roberts,  G  Gambino,  A  Cook,  T  Kouzarides,  and  C  R  Goding.  1997.  “CBP/p300  as  a  Co-­‐Factor  for  the  Microphthalmia  Transcription  Factor.”  Oncogene  14  (25):  3083–92.  doi:10.1038/sj.onc.1201298.  http://www.ncbi.nlm.nih.gov/pubmed/9223672.  

Scargle,  Jeffrey  D.  1982.  “Statistical  Aspects  of  Spectral  Analysis  of  Unevenly  Spaced  Data.”  The  Astrophysical  Journal  263:  835–53.  http://articles.adsabs.harvard.edu/cgi-­‐bin/nph-­‐iarticle_query?bibcode=1982ApJ...263..835S&db_key=AST&page_ind=0&data_type=GIF&type=SCREEN_VIEW&classic=YES.  

Schwienbacher,  C,  L  Gramantieri,  R  Scelfo,  a  Veronese,  G  a  Calin,  L  Bolondi,  C  M  Croce,  G  Barbanti-­‐Brodano,  and  M  Negrini.  2000.  “Gain  of  Imprinting  at  Chromosome  11p15:  A  Pathogenetic  Mechanism  Identified  in  Human  Hepatocarcinomas.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  97  (10):  5445–49.  doi:10.1073/pnas.090087497.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=25848&tool=pmcentrez&rendertype=abstract.  

Sedgewick,  Andrew  J,  Stephen  C  Benz,  Shahrooz  Rabizadeh,  Patrick  Soon-­‐Shiong,  and  Charles  J  Vaske.  2013.  “Learning  Subgroup-­‐Specific  Regulatory  Interactions  and  Regulator  Independence  with  PARADIGM.”  Bioinformatics  (Oxford,  England)  29  (13):  i62–70.  doi:10.1093/bioinformatics/btt229.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3694636&tool=pmcentrez&rendertype=abstract.  

Shechter,  Ishaiahu,  Peihua  Dai,  Liang  Huo,  and  Guimin  Guan.  2003.  “IDH1  Gene  Transcription  Is  Sterol  Regulated  and  Activated  by  SREBP-­‐1a  and  SREBP-­‐2  in  Human  Hepatoma  HepG2  Cells:  Evidence  That  IDH1  May  Regulate  Lipogenesis  in  Hepatic  Cells.”  Journal  of  Lipid  Research  44  (11):  2169–80.  doi:10.1194/jlr.M300285-­‐JLR200.  

Page 148: Rachel D. Melamed

137

Shrager,  J,  JM  Tenenbaum,  and  M  Travers.  2010.  “Cancer  Commons:  Biomedicine  in  the  Internet  Age.”  In  Collaborative  Computational  Technologies  for  Biomedical  Research,  edited  by  S  Elkins,  M  Hupcey,  and  A  Williams.  

Simka,  M.  2010.  “Seasonal  Variations  in  the  Onset  and  Healing  Rates  of  Venous  Leg  Ulcers.”  Phlebology  /  Venous  Forum  of  the  Royal  Society  of  Medicine  25  (1):  29–34.  doi:10.1258/phleb.2009.008072.  http://www.ncbi.nlm.nih.gov/pubmed/20118343.  

Smyth,  Gordon  K.  2004.  “Linear  Models  and  Empirical  Bayes  Methods  for  Assessing  Differential  Expression  in  Microarray  Experiments.”  Statistical  Applications  in  Genetics  and  Molecular  Biology  3  (January):  Article3.  doi:10.2202/1544-­‐6115.1027.  http://www.ncbi.nlm.nih.gov/pubmed/16646809.  

Sokal,  R.R.,  and  F.J.  Rohif.  1981.  Biometry:  The  Principles  and  Practice  of  Statistics  in  Biological  Research.  New  York:  Freeman.  

Stark,  Chris,  Bobby-­‐Joe  Breitkreutz,  Teresa  Reguly,  Lorrie  Boucher,  Ashton  Breitkreutz,  and  Mike  Tyers.  2006.  “BioGRID:  A  General  Repository  for  Interaction  Datasets.”  Nucleic  Acids  Research  34  (Database  issue):  D535–39.  doi:10.1093/nar/gkj109.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1347471&tool=pmcentrez&rendertype=abstract.  

Stein,  Brian  D,  Adriana  Bautista,  Glen  T  Schumock,  Todd  a  Lee,  Jeffery  T  Charbeneau,  Diane  S  Lauderdale,  Edward  T  Naureckas,  David  O  Meltzer,  and  Jerry  a  Krishnan.  2012.  “The  Validity  of  International  Classification  of  Diseases,  Ninth  Revision,  Clinical  Modification  Diagnosis  Codes  for  Identifying  Patients  Hospitalized  for  COPD  Exacerbations.”  Chest  141  (1):  87–93.  doi:10.1378/chest.11-­‐0024.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3251268&tool=pmcentrez&rendertype=abstract.  

Subramanian,  Aravind,  Pablo  Tamayo,  Vamsi  K  Mootha,  Sayan  Mukherjee,  Benjamin  L  Ebert,  Michael  A  Gillette,  Amanda  Paulovich,  et  al.  2005.  “Gene  Set  Enrichment  Analysis:  A  Knowledge-­‐Based  Approach  for  Interpreting  Genome-­‐Wide  Expression  Profiles.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  102  (43):  15545–50.  doi:10.1073/pnas.0506580102.  http://www.pnas.org/cgi/content/abstract/102/43/15545.  

Szczurek,  Ewa,  and  Niko  Beerenwinkel.  2014.  “Modeling  Mutual  Exclusivity  of  Cancer  Mutations.”  PLoS  Computational  Biology  10  (3):  e1003503.  doi:10.1371/journal.pcbi.1003503.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3967923&tool=pmcentrez&rendertype=abstract.  

Tang,  Ning,  John  Stein,  Renee  Y  Hsia,  Judith  H  Maselli,  and  Ralph  Gonzales.  2010.  “Trends  and  Characteristics  of  US  Emergency  Department  Visits,  1997-­‐2007.”  JAMA :  The  Journal  of  the  American  Medical  Association  304  (6).  American  Medical  Association:  664–70.  doi:10.1001/jama.2010.1112.  http://jama.jamanetwork.com/article.aspx?articleid=186383&resultClick=3.  

Page 149: Rachel D. Melamed

138

Taniguchi,  Kenichiro,  Anoush  E  Anderson,  Ann  E  Sutherland,  and  David  Wotton.  2012.  “Loss  of  Tgif  Function  Causes  Holoprosencephaly  by  Disrupting  the  SHH  Signaling  Pathway.”  PLoS  Genetics  8  (2):  e1002524.  doi:10.1371/journal.pgen.1002524.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3285584&tool=pmcentrez&rendertype=abstract.  

Tanno,  Toshihiko,  Natarajan  V  Bhanu,  Patricia  A  Oneal,  Sung-­‐Ho  Goh,  Pamela  Staker,  Y  Terry  Lee,  John  W  Moroney,  et  al.  2007.  “High  Levels  of  GDF15  in  Thalassemia  Suppress  Expression  of  the  Iron  Regulatory  Protein  Hepcidin.”  Nature  Medicine  13  (9).  Nature  Publishing  Group:  1096–1101.  doi:10.1038/nm1629.  http://dx.doi.org/10.1038/nm1629.  

Tarca,  Adi  Laurentiu,  Sorin  Draghici,  Purvesh  Khatri,  Sonia  S  Hassan,  Pooja  Mittal,  Jung-­‐Sun  Kim,  Chong  Jai  Kim,  Juan  Pedro  Kusanovic,  and  Roberto  Romero.  2009.  “A  Novel  Signaling  Pathway  Impact  Analysis.”  Bioinformatics  (Oxford,  England)  25  (1):  75–82.  doi:10.1093/bioinformatics/btn577.  http://bioinformatics.oxfordjournals.org/cgi/content/abstract/25/1/75.  

“The  Cancer  Genome  Atlas.”  Http://www.cancergenome.nih.gov/.  

Therneau,  Terry.  2012.  “A  Package  for  Survival  Analysis  in  S.  R  Package  Version.”  

Tieder,  Joel  S,  Matthew  Hall,  Katherine  a  Auger,  Paul  D  Hain,  Karen  E  Jerardi,  Angela  L  Myers,  Suraiya  S  Rahman,  Derek  J  Williams,  and  Samir  S  Shah.  2011.  “Accuracy  of  Administrative  Billing  Codes  to  Detect  Urinary  Tract  Infection  Hospitalizations.”  Pediatrics  128  (2):  323–30.  doi:10.1542/peds.2010-­‐2064.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3146355&tool=pmcentrez&rendertype=abstract.  

Torkamani,  Ali,  and  Nicholas  J  Schork.  2009.  “Identification  of  Rare  Cancer  Driver  Mutations  by  Network  Reconstruction.”  Genome  Research  19  (9):  1570–78.  doi:10.1101/gr.092833.109.  http://genome.cshlp.org/cgi/content/abstract/19/9/1570.  

Torti,  Suzy  V,  and  Frank  M  Torti.  2013.  “Iron  and  Cancer:  More  Ore  to  Be  Mined.”  Nature  Reviews.  Cancer  13  (5).  Nature  Publishing  Group,  a  division  of  Macmillan  Publishers  Limited.  All  Rights  Reserved.:  342–55.  doi:10.1038/nrc3495.  http://dx.doi.org/10.1038/nrc3495.  

Trifonov,  Vladimir,  Laura  Pasqualucci,  Enrico  Tiacci,  Brunangelo  Falini,  and  Raul  Rabadan.  2013.  “Statistical  Algorithm  for  Variant  Frequency  Identification.”  BMC  Systems  Biology  In  Press.  

Turcan,  Sevin,  Daniel  Rohle,  Anuj  Goenka,  Logan  A.  Walsh,  Fang  Fang,  Emrullah  Yilmaz,  Carl  Campos,  et  al.  2012.  “IDH1  Mutation  Is  Sufficient  to  Establish  the  Glioma  Hypermethylator  Phenotype.”  Nature.  doi:10.1038/nature10866.  

Page 150: Rachel D. Melamed

139

Upshur,  Ross  E  G,  Rahim  Moineddin,  Eric  Crighton,  Lori  Kiefer,  and  Muhammad  Mamdani.  2005.  “Simplicity  within  Complexity:  Seasonality  and  Predictability  of  Hospital  Admissions  in  the  Province  of  Ontario  1988-­‐2001,  a  Population-­‐Based  Analysis.”  BMC  Health  Services  Research  5  (1):  13.  doi:10.1186/1472-­‐6963-­‐5-­‐13.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=549216&tool=pmcentrez&rendertype=abstract.  

Vandin,  Fabio,  Eli  Upfal,  and  Benjamin  J  Raphael.  2011.  “Algorithms  for  Detecting  Significantly  Mutated  Pathways  in  Cancer.”  Journal  of  Computational  Biology :  A  Journal  of  Computational  Molecular  Cell  Biology  18  (3):  507–22.  doi:10.1089/cmb.2010.0265.  http://www.ncbi.nlm.nih.gov/pubmed/21385051.  

———.  2012.  “De  Novo  Discovery  of  Mutated  Driver  Pathways  in  Cancer.”  Genome  Research  22  (2):  375–85.  doi:10.1101/gr.120477.111.  http://www.ncbi.nlm.nih.gov/pubmed/21653252.  

Varadan,  Vinay,  and  Dimitris  Anastassiou.  2006.  “Inference  of  Disease-­‐Related  Molecular  Logic  from  Systems-­‐Based  Microarray  Analysis.”  PLoS  Computational  Biology  2  (6):  e68.  doi:10.1371/journal.pcbi.0020068.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1479089&tool=pmcentrez&rendertype=abstract.  

Vaske,  Charles  J,  Stephen  C  Benz,  J  Zachary  Sanborn,  Dent  Earl,  Christopher  Szeto,  Jingchun  Zhu,  David  Haussler,  and  Joshua  M  Stuart.  2010.  “Inference  of  Patient-­‐Specific  Pathway  Activities  from  Multi-­‐Dimensional  Cancer  Genomics  Data  Using  PARADIGM.”  Bioinformatics  (Oxford,  England)  26  (12):  i237–45.  doi:10.1093/bioinformatics/btq182.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2881367&tool=pmcentrez&rendertype=abstract.  

Verhaak,  Roel  G  W,  Katherine  a  Hoadley,  Elizabeth  Purdom,  Victoria  Wang,  Yuan  Qi,  Matthew  D  Wilkerson,  C  Ryan  Miller,  et  al.  2010.  “Integrated  Genomic  Analysis  Identifies  Clinically  Relevant  Subtypes  of  Glioblastoma  Characterized  by  Abnormalities  in  PDGFRA,  IDH1,  EGFR,  and  NF1.”  Cancer  Cell  17  (1).  Elsevier  Ltd:  98–110.  doi:10.1016/j.ccr.2009.12.020.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2818769&tool=pmcentrez&rendertype=abstract.  

Vlachos,  Adrianna,  Philip  S  Rosenberg,  Eva  Atsidaftos,  Blanche  P  Alter,  and  Jeffrey  M  Lipton.  2012.  “Incidence  of  Neoplasia  in  Diamond  Blackfan  Anemia:  A  Report  from  the  Diamond  Blackfan  Anemia  Registry.”  Blood  119  (16):  3815–19.  doi:10.1182/blood-­‐2011-­‐08-­‐375972.  http://bloodjournal.hematologylibrary.org/content/119/16/3815?variant=long&sso-­‐checked=1.  

Wang,  Ping-­‐Yuan,  Jian  Weng,  and  Richard  G  W  Anderson.  2005.  “OSBP  Is  a  Cholesterol-­‐Regulated  Scaffolding  Protein  in  Control  of  ERK  1/2  Activation.”  Science  (New  York,  N.Y.)  307  (December  2004):  1472–76.  doi:10.1126/science.1107710.  

Page 151: Rachel D. Melamed

140

Wang,  Xiaoxia,  Sherry  Towers,  Sarada  Panchanathan,  and  Gerardo  Chowell.  2013.  “A  Population  Based  Study  of  Seasonality  of  Skin  and  Soft  Tissue  Infections:  Implications  for  the  Spread  of  CA-­‐MRSA.”  PloS  One  8  (4):  e60872.  doi:10.1371/journal.pone.0060872.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3614932&tool=pmcentrez&rendertype=abstract.  

Wang,  Yingli,  Bo  Wan,  Dawei  Li,  Jun  Zhou,  Ruwei  Li,  Meirong  Bai,  Fang  Chen,  and  Long  Yu.  2012.  “BRSK2  Is  Regulated  by  ER  Stress  in  Protein  Level  and  Involved  in  ER  Stress-­‐Induced  Apoptosis.”  Biochemical  and  Biophysical  Research  Communications  423  (4):  813–18.  doi:10.1016/j.bbrc.2012.06.046.  http://www.sciencedirect.com/science/article/pii/S0006291X12011345.  

Watanabe,  Satosi.  1960.  “Information  Theoretical  Analysis  of  Multivariate  Correlation.”  IBM  Journal  of  Research  and  Development  4  (January):  66–82.  doi:10.1147/rd.41.0066.  

Watson,  Ian  R,  Koichi  Takahashi,  P  Andrew  Futreal,  and  Lynda  Chin.  2013.  “Emerging  Patterns  of  Somatic  Mutations  in  Cancer.”  Nature  Reviews.  Genetics  14  (10).  Nature  Publishing  Group:  703–18.  doi:10.1038/nrg3539.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=4014352&tool=pmcentrez&rendertype=abstract.  

Welcker,  Markus,  Amir  Orian,  Jianping  Jin,  Jonathan  E  Grim,  J  Wade  Harper,  Robert  N  Eisenman,  and  Bruce  E  Clurman.  2004.  “The  Fbw7  Tumor  Suppressor  Regulates  Glycogen  Synthase  Kinase  3  Phosphorylation-­‐Dependent  c-­‐Myc  Protein  Degradation.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  101  (24):  9085–90.  

Wilkerson,  Matthew  D,  and  D  Neil  Hayes.  2010.  “ConsensusClusterPlus:  A  Class  Discovery  Tool  with  Confidence  Assessments  and  Item  Tracking.”  Bioinformatics  (Oxford,  England)  26  (12):  1572–73.  doi:10.1093/bioinformatics/btq170.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2881355&tool=pmcentrez&rendertype=abstract.  

Winthorst,  Wim  H,  Wendy  J  Post,  Ybe  Meesters,  Brenda  W  H  J  Penninx,  and  Willem  A  Nolen.  2011.  “Seasonality  in  Depressive  and  Anxiety  Symptoms  among  Primary  Care  Patients  and  in  Patients  with  Depressive  and  Anxiety  Disorders;  Results  from  the  Netherlands  Study  of  Depression  and  Anxiety.”  BMC  Psychiatry  11  (January):  198.  doi:10.1186/1471-­‐244X-­‐11-­‐198.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3280179&tool=pmcentrez&rendertype=abstract.  

Wu,  Di,  and  Gordon  K  Smyth.  2012.  “Camera:  A  Competitive  Gene  Set  Test  Accounting  for  Inter-­‐Gene  Correlation.”  Nucleic  Acids  Research  40  (17):  e133.  doi:10.1093/nar/gks461.  http://nar.oxfordjournals.org/content/40/17/e133.  

Wu,  Guanming,  Xin  Feng,  and  Lincoln  Stein.  2010.  “A  Human  Functional  Protein  Interaction  Network  and  Its  Application  to  Cancer  Data  Analysis.”  Genome  Biology  11  (5):  R53.  

Page 152: Rachel D. Melamed

141

doi:10.1186/gb-­‐2010-­‐11-­‐5-­‐r53.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2898064&tool=pmcentrez&rendertype=abstract.  

Yajima,  Ichiro,  Mayuko  Y  Kumasaka,  Nguyen  Dinh  Thang,  Yuji  Goto,  Kozue  Takeda,  Machiko  Iida,  Nobutaka  Ohgami,  et  al.  2011.  “Molecular  Network  Associated  with  MITF  in  Skin  Melanoma  Development  and  Progression.”  Journal  of  Skin  Cancer  2011  (January):  730170.  doi:10.1155/2011/730170.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3199194&tool=pmcentrez&rendertype=abstract.  

Yamamoto,  Yukiya,  Akihiro  Abe,  and  Nobuhiko  Emi.  2014.  “Clarifying  the  Impact  of  Polycomb  Complex  Component  Disruption  in  Human  Cancers.”  Molecular  Cancer  Research :  MCR  12  (4):  479–84.  doi:10.1158/1541-­‐7786.MCR-­‐13-­‐0596.  http://mcr.aacrjournals.org/content/12/4/479.long#ref-­‐16.  

Yang,  Xiang-­‐Jiao.  2004.  “The  Diverse  Superfamily  of  Lysine  Acetyltransferases  and  Their  Roles  in  Leukemia  and  Other  Diseases.”  Nucleic  Acids  Research  32  (3):  959–76.  doi:10.1093/nar/gkh252.  http://nar.oxfordjournals.org/content/32/3/959.full.  

Zheng,  Yu,  Zebin  Wang,  Wenjun  Bie,  Patrick  M  Brauer,  Bethany  E  Perez  White,  Jing  Li,  Veronique  Nogueira,  et  al.  2013.  “PTK6  Activation  at  the  Membrane  Regulates  Epithelial-­‐Mesenchymal  Transition  in  Prostate  Cancer.”  Cancer  Research  73  (17).  American  Association  for  Cancer  Research:  5426–37.  doi:10.1158/0008-­‐5472.CAN-­‐13-­‐0443.  http://cancerres.aacrjournals.org/content/73/17/5426.full.  

Zhu,  Haihao,  Jaime  Acquaviva,  Pranatartiharan  Ramachandran,  Abraham  Boskovitz,  Steve  Woolfenden,  Rolf  Pfannl,  Roderick  T  Bronson,  et  al.  2009.  “Oncogenic  EGFR  Signaling  Cooperates  with  Loss  of  Tumor  Suppressor  Gene  Functions  in  Gliomagenesis.”  Proceedings  of  the  National  Academy  of  Sciences  of  the  United  States  of  America  106  (8):  2712–16.  doi:10.1073/pnas.0813314106.  http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2650331&tool=pmcentrez&rendertype=abstract.