a b Copy Number Alterations in M-class TP53 Mutation Type · PDF file10 20 30 40 50 60 TP53 Mutation Type 70 ... of which thyroid samples are a subset. d, The ranked list of SFEs,

TP53

-mut

ated

sam

ples

(%)

M-class C-class

MissenseTruncating

1:2.45

1:3.85

0

10

20

30

40

50

60

TP53 Mutation Type

70

Copy Number Alterations in M-class

TP53 WT TP53 mut

0

2

4

6

8

Rec

urre

nt C

opy

Num

ber A

ltera

tions

a b

Supplementary Figure 1

Supplementary Figure 1: TP53-mutated tumors a, TP53-mutated tumors within the M-class have a higher number of recurrent copy number alterations (for all SFEs) than TP53 wild type samples. b, While both missense (residue-changing) and truncating TP53 mutations are more frequent in the C-class than in the M-class, this enrichment is significantly stronger for truncating mutations.

Nature Genetics: doi:10.1038/ng.2762

No. of recurrent CNAs

No.

of r

ecur

rent

mut

atio

ns

OnlyTumor fromMutated class

Only tumors fromCN-Altered class

Both are present

Ratio of tumors fromthe two classes


0 10 20 30 40

010

2030

40

No. of CNA

No.

of m

utat

ions

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40

0 10 20 30 40

010

2030

40No. of CNA No. of CNA No. of CNA

No.

of m

utat

ions

No.

of m

utat

ions

No.

of m

utat

ions

No. of CNA

No.

of m

utat

ions

No. of CNA No. of CNA No. of CNA

No.

of m

utat

ions

No.

of m

utat

ions

No.

of m

utat

ions

No. of CNA

No.

of m

utat

ions

No. of CNA No. of CNA No. of CNA

No.

of m

utat

ions

No.

of m

utat

ions

No.

of m

utat

ions

ALL TUMORS BLCA BRCA COADREAD

GBM HNSC KIRC LAML

LUAD LUSC OV UCEC


Supplementary Figure 2: Mutations versus copy number alterations. Tumors with more recurrent mutations than copy number alterations predominantly are in the M-class (increasing intensity of green), while tumors with more recurrent copy number alterations mostly are in the C-class (increasing intensity of red). Similar levels of both type of alterations are in tumors at the border between the two classes (grey).

Supplementary Figure 3: Distribution of mutations and copy number changes across different tumor types. Samples from different tumor types and subtypes vary in the number of recurrent copy number alterations (X-axis) and number of recurrent mutations (Y-axis).


ProstateAdenocarcinoma(170)

PancreasAdenocarcinoma

(27)

Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma(35)

Thyroid PapillaryCarcinoma

(400)

StomachAdenocarcinoma(219)

Skin CutaneousMelanoma (58)

Independent TCGA Dataset (ITD) PANCAN18 Hyperbola

40

30

20

10

0

0 10 20 30 40

PANCAN12ITD

BRAF-mutantThyroid-enriched

class

PANCAN18

C-class

M-class

(378) (1734)

(2069)

150

100

500

5010

0

High Amp.Hom. Del.Mutations

# of

Rec

urre

nt M

utat

ions

# of Recurrent Copy Number Alterations

Enric

hmen

t [-lo

g(q)

]

M-class vs C-class (PANCAN18)

KRAS

PIK3CA

APC

ARID1A

TP53 (q = 8E-219)

8q24 (MYC)

3q26

Alterations Enriched in M-class

Alterations Enriched in C-class

τ = 0.9*

a b

c d

Characterized by somatic mutations

Characterized by CNAs


Supplementary Figure 4: Robustness of classes and features - validation in enlarged dataset. a, An independent dataset consisting of 909 samples from 6 tumor types was integrated with our original dataset (PANCAN12). b, Samples from the resulting set of 18 tumor types (PANCAN18) confirmed the inverse relationships between copy number alterations and mutations, with the newly added samples following the same inverse trend (red squares) we originally observed (gray squares). c, A stratification of the PANCAN18 dataset identifies two main modules of similar size and a small subgroups enriched in BRAF-mutant thyroid samples. These thyroid tumors emerged separately as strongly characterized by BRAF V600E mutations and otherwise depleted of any other SFE. Nonetheless, these tumors do not violate the distinction between tumors primarily characterized by copy number alterations (C-class) and tumors characterized by mutations (M-class), of which thyroid samples are a subset. d, The ranked list of SFEs, with the first of the list being the most enriched in the M-class and the last the most enriched in C-class, is almost perfectly concordant with the same ranked list obtained in the PANCAN12 dataset (Kendall correlation coefficient tau = 0.9).


ALLTUMORS

M1M2M3M4M5M6M7M8

M9M10M11M12M13M14

M15

M16

M17

C7

C8C9

C10C11C12

C13

C14

C1C2C3C4C5C6

M

Mutations on APC (50%), PTEN (47%), POLE (46%), ATM (43%) MLH1-methylation (71%), BRAF mutations (57%), 16p deletion (46%)ARID1A (91%) and CTNNB1 (32%) mutationsPTEN (94%) and PIK3R1 (31%) mutationsPTEN (79%), PIK3CA (64%), ARID1A (61%), CTCF (40%), CTNNB1 (31%) mutations and MLH1 methylation (46%) PIK3CA (99%) mutationsPIK3CA (96%), KRAS (72%), APC (53%) mutations and MGMT (46%) methylationPIK3CA (97%) and TP53 (75%) mutations

APC (76%), TP53 (57%) mutations and 20q amplification (84%)APC (88%) and NRAS (19%) mutationsTP53 (93%), APC (83%) mutations and MGMT (34%) methylationAPC (86%), KRAS (86%), and TP53 (52%) mutationsKRAS (99%), TP53 (45%), APC (40%) mutations and MGMT (48%) methylationMGMT methylation (84%), TP53 mutations (56%), and 9q21 (CDKN2A) deletion (35%)

9q21 (CDKN2A) deletion (76%) and 7p11 (EGFR) amplification (48%)

FLT3 (38%), NPM1 (29%), and DNMT3A (27%) mutations

VHL (63%), PBRM1 (41%) mutations and GSTP1 methylation (36%)

TP53 mutations (76%), 8p21 (PPP2R2A* - 56%), 8p23 (FBXO25* - 48%) deletion, and 1q21 (SETDB1* - 40%), 8q24 (MYC - 36%) and 3q26 (ZNF639*, PIK3CA - 35%) amplification

TP53 mutations (72%) and 8p23 (FBXO25* - 81%), 8p23 (PPP2R2A* - 64%) deletion8q24 (MYC - 93%), 8q22 (YWHAZ* - 91%), 8p11 (IKBKB - 55%), 8p12 (WHSC1L1, FGFR1 - 45%) amplification and TP53 mutations (36%)TP53 mutations (100%) and 8q24 (MYC - 95%), 8q22 (YWHAZ* - 100%) amplification8q24 (MYC) amplification (99%), TP53 (89%) and BRCA2 (25%) mutations17q23-25 (RPS6KB1, mir-21 - 67%), 17q21 (ERBB2 - 28%), 8q24 (MYC - 56%), 8q22 (YWHAZ* - 42%), 1q21 (SETDB1* - 35%), 1q34 (PPP1R15B*, MDM4 - 33%) amplification and TP53 mutations (36%)6p22-23 (E2F3 - 88%), 8q24 (MYC - 71%), 3q26 (ZNF639*, PIK3CA - 50%), 8q22 (YWHAZ* - 38%), 7q36 (UBE3C* 33%) amplification and TP53 (75%), BRCA2 (31%), BRCA1 (29%) mutations8q24 (MYC - 88%), 3q26 (ZNF639* - 78%), 20q13 (AURKA - 52%), 8q22 (YWHAZ* - 50%), 1q21 (SETDB1* - 44%), 5p15 (CLPTM1L* - 33%), 19p13 (BRD4* - 27%) amplification and TP53 mutations (78%)

TP53 mutations (99%)TP53 mutations (96%) and 19q12 (CCNE1 - 40%), 19p13 (BRD4* - 32%), 3q26 (ZNF639*, PIK3CA - 31%) amplificationTP53 mutations (92%), 3q26 (ZNF639*, PIK3CA) amplification (64%) and 9q21 (CDKN2A) deletion (32%) TP53 (84%), CDKN2A (34%) mutations, 11q13 (CCND1) amplification (82%), 9q21 (CDKN2A) deletion (25%)TP53 mutations (53%), 11q13 (CCND1 - 79%), 11q14 (PAK1 - 38%), 8p11 (WHSC1L1, FGFR1 - 30%) amplification1q21 (SETDB1* - 81%), 1q34 (PPP1R15B*, MDM4 - 56%) amplification and TP53 mutations (54%)

Ultra mutators UCEC and COADREAD (57-61%)Micro-satellite instable COADREAD (56-79%) and UCEC (13-16%)Mixed (BLCA - 20%, LUAD - 15%, UCEC low CNA - 15%)Mixed (GBM - 30%, UCEC Low CNA - 24%)UCEC MSI (43%), UCEC Low CNA (33%), BRCA Lum A (11%)BRCA Lum A (48%), HNSC (14%), KIRC (10%)COADREAD MSS (40%) and MSI (12%), UCEC Low CNA (10%)Mixed (HNSC - 25%, COADREAD MSS - 13%, GBM - 13%, BLCA -12%)

COADREAD MSS (56-79%), LUAD (10%)COADREAD MSS (53-78%)COADREAD MSS (48-80%)COADREAD MSS (48-85%)COADREAD MSS (37-56%), LUAD (34%)GBM (41%), AML (16%), HNSC (13%)

GBM (61%), LUAD (12%), HNSC (11%)

AML (90%)

KIRC (92%)

OV (56%), BRCA Basal (15%)

Mixed (HNSC - 20%, OV - 16%, BRCA Lum A - 15%, LUSC - 13%)BRCA Lum A (33%) and Lum B (15%), BLCA (7%)OV (40%), HNSC (21%), BRCA Basal (10%)OV (50%), UCEC Serous (8%), BRCA Basal (8%), LUAD (8%)BRCA Lum A (21%), LumB (19%), Her2+ (10%), Basal (9%),OV (18%)OV (75%), BRCA Basal (8%)

OV (68%), BRCA Basal (9%)

Mixed (HNSC -20%, OV - 20%, LUAD - 15%)OV (66%), LUSC (11%), UCEC Serous (6%)LUSC (40%), HNSC (25%)HNSC (50%), LUSC (16%)HNSC (30%), BRCA Lum A (25%) and Lum B (16%)Mixed (BRCA Lum A - 19%, OV - 16%, LUAD - 15%, UCEC Serous - 11%)

CLUSTERSIZE

46

56

34

135

61

99

57

101

62

32

145

65

147

108

126

115

350

191

128

163

44

122

138

75

100

55

58

75

78

48

110

C

TUMOUR HIERARCHICALCLASSIFICATION MOST FREQUENT SFEs MOST REPRESENTED TUMOR TYPES


* Most differentially expressed gene in the region when compared to diploid samples

Supplementary Figure 5: Hierarchical stratification. The hierarchical stratification of tumors based on selected functional events identifies 31 oncogenic signature sub-classes (dendrogram on the left). For each sub-class, we report the most frequent SFEs and the most represented tumor types. For each recurrent copy number alteration, we report known oncogenes or tumor suppressors in the region (if present). For large regions spanning multiple genes, we also report the gene with the most significant differential mRNA expression between altered and diploid cases.


Jacc

ard

Coe

ffici

ent

Jacc

ard

Coe

ffici

ent

1st-l

evel

m

odul

es (5

%)

1st-l

evel

m

odul

es (2

0%)

2nd-

leve

l m

odul

es (5

%)

2nd-

leve

l m

odul

es (2

0%)

M1-

M17

(5%

)

M1-

M17

(20%

)

2nd-

leve

l m

odul

es (5

%)

2nd-

leve

l m

odul

es (2

0%)

C1-

C14

(5%

)

C1-

C14

(20%

)

M subclasses C subclasses Random Classification

ALLTUMORS

M C M/C classes

1st-level modules

2nd-level modules

1st-l

evel

m

odul

es (5

%)

1st-l

evel

m

odul

es (2

0%)

M1 M2 M3 M4 M5 M6 M7 M8 M9M10M11M12M13M14 M15 M16 M17 C1 C2 C3 C4 C5 C6 C7C8C9C10C11C12C13C14

1795 samples 1445 samples

0

0.2

0.4

0.6

0.8

1

0

0.2

0.4

0.6

0.8

1

0

0.2

0.4

0.6

0.8

1

Top

Cla

sses

(5

%)

Top

Cla

sses

(2

0%)

Top

Cla

sses

(5

0%)

Jacc

ard

Coe

ffici

ent

M/C-classed


Supplementary Figure 6: Robustness of cluster assignments. Subclasses at all levels of the hierarchical stratification were separately tested for robustness to random sample removal. The top classes M and C (black dots) show almost perfect reproducibility upon removal of 5%, 20%, and 50% of the samples. Similarly, the M subclasses (green dots) maintain a constantly high robustness descending the classification. C subclasses (red dots) at the bottom the classification are the most affected by sample removals, despite a robustness significantly higher than expected (gray dots). The non-focal nature of copy number alterations and the difficulty in identifying the corresponding functional targets are likely to impact the definition of these groups, weakening their robustness.


ARID1A Mutated Samples (%)

100%

50%

0

ARID1A Hotspot Mutations

R1989* (x14)

R1721* (x2)R1722* (x5)

Q1334 IF-Del (x4)Q1334 IF-Ins (x2)R1335* (x5)

R693* (x5)

0 2285POLE / ultra-mutators


M5M3M1

a b

R283CH284N/Y/P

G261-splice (x3)

H312RN314fs (x3)

R377C (x2)R377H (x2)P378L (x2)P378A R448* (x7)

CTCF Hotspot Mutations

0 727

EX 3 EX 5EX 4

EX 3 EX 5

Wild-type

G261-splice (TCGA-A2-AY0F)

EX 4

Exon Skipping

ZF1 ZF2

CTCF Splice-site Mutation

His-1(H284/H312)

His-2Cys-2

Cys-1

Zn

CTCF C2-H2 Zinc Finger

50%

0

25%

c d

e f

Zinc-Finger domains (ZF1-ZF11)

M5

BR

CA

UC

EC

KIRCHNSC

CTCF Mutated Samples (%)

Supplementary Figure 7: Recurrent mutations identified by cross-cancer analysis. a, ARID1A mutations are observed in sub-classes M1, M3, and M5, which include diverse tumor subtypes (colored stacked bars). b, The frequency of mutations of ARID1A across all tumors indicates hotspot regions (colored according to the tumor type of the mutated samples). c, CTCF mutations are characteristic of sub-class M5 and d, hotspot somatic mutations target different zinc-finger domains (white boxes). e, hotspot mutations frequently occur at the zinc-binding histidine. f, This splice-site mutation causes an in-frame exon-skipping event. The skipped exon codes for zinc-finger 1 and 2.


67,650 kb

26 kb

p12.3 p12.2 p12.1 p11.2 p11.1 q11.2 q12.1 q12.2 q13 q21 q22.1 q22.2

Chr. 16

p13.3 p13.2p13.13 p13.11 p12.3 p12.2 p12.1 p11.2 p11.1 q11.2 q12.1 q12.2 q13 q21 q22.1 q22.2 q23.1 q23.2 q23.3 q24.1q24.2

CTCF

CTCF

Exon 3Exon 4

Exon 5

Exon4 Skipping

TCGA-A2-AY0F


Exon 6 Exon 7

Supplementary Figure 8: Splice-site mutation in CTCF. Splice-site mutation on G261 on CTCF (located on 16q22) caused the skipping of Exon 4 in the TCGA-A2-AY0F sample. Multiple RNA-seq reads (grey thick lines, one sample per row) span sequences starting from Exon 3 to Exon 5, therefore skipping Exon 4 (endpoints of reads spanning more than one exon are connected by blue lines).


1q21 1q32 3q26 8p23 8p21 19p13

050

100

150

200

com

bine

d q-

valu

es

PPP1R15BMDM4

ZNF639MFN1PIK3CA

FBXO25PPP2R2A

BRD4

1q AMP 3q AMP 8p DEL 19p AMP

SETDB1

Supplementary Figure 9G

ISTI

Cq-

valu

es

Supplementary Figure 9: Significant loci within chromosomal events. Recurrent gains and losses on chromosomes 1q, 3q, 8p, and 19p span multiple genes. The Gistic algorithm identified the most significant loci in these regions (peaks of q-value histograms). We tested each gene in these peaks for differential mRNA expression concordant with the copy number change (over-expressed if amplified, down-regulated if deleted, relative to diploid cases). Combined q-values [-log(q)] serve to determine the most likely functional targets (labeled by corresponding gene symbol).


CCND1

E280delT286IP287S

1234567

295 aa

# M

utat

ions

00

20

40

60

80

100

120

140

# M

utat

ions

PIK3CA 1068 aa

PESTCyclin Box

H1047L/Q/R/YE545A/D/G/K/Q

H542A/G/K/Q/V

R88QQ546E/K/P/RN345I/K/T

M1043I/VE726KG118D

a

b


Supplementary Figure 10: Hotspot mutations identified by cross-cancer analysis. a, PIK3CA hotspot mutations. b, Hotspot mutations in CCND1 cluster in the PEST domain of the protein, preventing degradation and stabilizing the protein. Mutations are indicated by vertical bars (bar height proportional to the number of mutations in the specified residue, summed over all affected samples in all tumor types).


99 96 94 84 59 89 89 86 52 100 91 72 85 85 59 26 28 5 5 4 77 54 7 94 52 52 60 10 3 60

99 96 92 84 53 54 76 72 36 100 89 36 75 78 22 11 12 21 2 75 60 3 93 52 44 56 501 2 6 54 20 2 2 1 29 6 17 3 3 1 2 2 1 2 3 5 53 3 7 5 2 5 5 18 10 5 8 10 2 6 3 3 2 2 2 4 2 2 6 4

1 1 2 2 1 1 8 4 43 17 9 3 2 2 3 4 2 5 7 2 1 2 44 17 4 2 1 1 15 4 8 5 16 14 27 22 34 8 2 2 1 1 62 11 2 1 2 16 7 2 9 11 7 25 13 9 2 2 1 2 41 4 2 1 1 16 3 5 7 8 25 18 3

30 62 68 84 86 51 68 63 44 40 49 53 75 73 30 6 29 46 12 17 4 63 54 12 18 2 13 53 82 8 42

3 13 7 2 8 11 19 16 10 12 4 22 17 51 3 2 4 2 4 54 5 1 91 4 1 1 2 27 4 3 18 69 14 1 2 2 4 4 58 5 7 3 4 3 1 3 2 2 16 2 4 2 2 5 1 2 7 3 33 3 5 3 4 9 4 2 4 5 3 10 6 2 2 3 2 1 4 1 3 2 3

40 1 2 5 17 26 5 6 2 23 5 8 21 3 2 53 1 2 2 3 6 6 1 4 2 4 8 3 2 4 2 2 6 6 21 1 7 5 7 1 1 2 4 1 1 3 1 1 1 2 1 1

1 2 2 1 3 1 11 1 81 79 10 9 20 21 19 4 19 4 11 6 3 2 8 2 6 2 6 2 10

1 1 1 7 6 2 1 1 4 2 1 1 11 2 1 1 3 1 1 4 1

12 1 8 28 7 1 1 6 4 4 2 3 2 12 5 2 4 1 32 3 12 2 3 2 1 1 4 2 1 1

6 32 23 17 12 4 28 10 7 5 10 3 15 26 31 2 4 1 6 35 75 6 12

10 44 71 23 39 51 57 20 31 21 36 32 50 69 68 35 35 98 83 99 96 98 4 9 5 20 18 14 2 6 40

1 1 25 4 3 2 2 1 2 1 1 1 1 11 1 2 1 1 1 1 1 2 2 3 1 2 2 2 1 2 1 2 1 1

2 1 1 11 1 2 1 2 3 1 1 3 1 1 2 10 1 1 11 9 1 3 1 23 1 2 2 1 2 3 2 2 2 1 1 2

1 1 4 2 1 1 2 2 12 2 4 3 4 1 3 3 4 1 48 8 9 94 78 6 8 5 94 3 6 3 9 4 2 5 4 8 2 6 4 3 7 1 2 2 5 8 1 32 1 2 3 3 2 2 1 1 36 10 9 30 17 2 2 1 2 6 2 1 42 5 10 2 12 13 3 4 23 5 9 5 3 32 10 12 2 63 99 96 97 2 1 3 17

31 64 16 24 30 35 14 4 9 16 16 50 57 2 6 2 4 5 1 2 1 6 1 2 13

29 38 52 9 50 51 43 56 63 28 36 53 35 54 50 69 29 33 44 8 86 38 48 31 33 82 100 44 66 10 6 44

3 1 1 3 4 1 1 9 60 2 3 6 1 2 9 3 2 1 42 1 2 2 1 5 6 2 5 4 2 15 8 2 6 2 10 2

2 2 4 3 2 1 2 4 2 3 1 11 4 3 2 2 1 1 16 12 6 27 76 12 17 3 82 99 1 1 12

1 17 3 2 4 9 16 9 5 7 8 15 11 2 1 4 4 3 1 44 2 4 1 3 4 4 4 1 6 5 30 4 6 2 4 2 3 1 4 2 1 32 4 1 3 12 2 2 2 3 1 2 6 2 11 1 5 2 1 3 6 3 4 4 4 1 2 13 3 2

7 1 2 1 3 2 3 1 13 13 1 2 1 2 3 1 21 1 1 2 1 3 1 11 2 1 4 7 3 4 4 8 8 13 1 2 2 2 21 1 1 3 1 2 2 11 6 3 5 7 1 1

6 26 29 6 3 24 46 7 6 3 2 10 8 3 2 1 1 61 1 2 2 2 6 1 3 1 4 2 5 2 1 1 1

1 6 1 1 7 4 1 4 9 4 1 11 1 1 1 3 2 1 1 4 2 2 3 1 2 15 4 7 9 13 4 5 15 5 8 30 4 9 2 3 1 2 2 4 10 2 2 55 4 1 2 5 2 1 1 2 2 2 3 7 18 1 28

C1 C2

10

C3

7

C4

8

C5

7

C6

1

C7

9

C8

4

C9

4

C10

7

C11

4

C12

2

C13

2

C14 M1 M2 M3

13

M4 M5 M6

2

M7

8

M8 M9 M10

3

M11 M12

1

M13

14

M14

48

M15 M16

1

M17

5

Averag

e

p53/DNA-repairTP53

MDM4MDM2

ATMBRCA2BRCA1BRCA1

Cell−Cycle

AURKAE2F3RB1RB1

CCNE1CDK4CDK6

CCND1CCND1

CDKN1BCDKN1ACDKN2ACDKN2ACDKN2A

PI(3)K/AKT/MTOR

MTORTSC1/2

TSC1STK11STK11

AKT1PTENPTEN

PIK3R1PIK3CAPIK3CA

RTK/Ras/RafBRAFNRASHRASKRASKRAS

NF1NF1

PDGFRA/KIT/KDRMET

FGFR3FGFR3FGFR2FGFR1ERBB3ERBB3ERBB2ERBB2EGFREGFR

Drug fa

mily

Pathway

RTK

inhi

bitio

nB

RA

F/M

AP

Ki

RTKs

Ras

Raf

PI(3)K

AktSTK11

TSC-complex

p21/p27

Rb

E2F-3AURK

Mdm2/4

p53

Cyclin D -CDK4/6

ATM

p16-INK4a

PTEN

NF1

PI(3

)Ki

AK

TiC

DK

4/6i

AU

RK

iM

DM

i

Cyclin E

mTOR

mTO

Ri

CD

K2i

BRCA1/2

PA

RP

i

ActivatingInactivating # pathways altered

0 1 2 3% altered

8 4DNA hyper-methylationHigh-level amplification

Somatic mutationHomozygous deletion

Currently targetable

DirectNo Indirect

100

50

0Sam

ples

(%)

33


Supplementary Figure 11: Map of functional and actionable alterations across 31 tumor subclasses. Four major oncogenic pathways (RTK/Ras/Raf, PI(3)K/Akt/mTOR, Rb, and DNA-repair/p53 signaling – simplified in the “Pathway” column) are altered by multiple different mechanisms (arranged vertically) across tumor types and subtypes (from left to right). Alterations to at least one of these pathways are observed in almost all samples of almost all tumor types (stacked barplots at the bottom). Several of these alterations are directly or indirectly therapeutically actionable (“Drug Class” column).



Event

Woman

Subclass A Subclass B

Supplementary Figure 12: Exercise to test algorithm: network subclasses identified in the Southern Women Club Attendance network. We identified two main classes (blue and red) and within each main class we identified smaller groups of women, who attended the same set of events. Grouping by dotted lines. These results are consistent with the ones in the original publications (see Methods).


Documents

a b Copy Number Alterations in M-class TP53 Mutation Type · PDF file10 20 30 40 50 60 TP53 Mutation Type 70 ... of which thyroid samples are a subset. d, The ranked list of SFEs,