CS177 Lecture 10 SNPs and Human Genetic Variation

Tom Madej 11.21.05

Lecture overview

• Human genetic variation, HapMap project.

• Experimental methods: PCR, X-ray crystallography, microarrays.

Motivations to study human genetic variation

• The evolution of our species and its history.

• Understand the genetics of diseases, esp. the more common complex ones such as diabetes, cancer, cardiovascular, and neurodegenerative.

• To allow pharmaceutical treatments to be tailored to individuals (adverse reactions based on genetics).

Genetic variation

• The human genome has approximately 10 million polymorphisms, i.e. genetic variants that occur at the level of about 1% or more in the population.

• Many of these polymorphisms are SNPs, single nucleotide polymorphisms.

• These polymorphisms contribute to our individuality, and also influence our susceptibility to various diseases.

Mendelian and non-Mendelian diseases

• Geneticists have been very successful in discovering the variations due to Mendelian disorders. These are characterized by in that they follow the Mendelian rules of inheritance.

• The study of particular families using linkage analysis has been successful for the Mendelian diseases.

• However, the more common complex (i.e. non-Mendelian) disorders have been much more difficult to investigate, even there there are clearly genetic components to many of these diseases.

Sources of genetic variation (during meiosis)

• Chromosomal reassortment; a human has 23 pairs of chromosomes, one of each pair is inherited from the father, and the other one from the mother.

• Mutation; errors in DNA copying. This may result in SNPs or also larger portions of DNA may be duplicated or copied incorrectly.

• Genetic recombination; shuffling of segments between partner chromosomes of a pair.

Reassortment of genetic material during meiosis

Molecular Biology of the Cell, Alberts et al. Garland Publishing 2002 (Fig. 20-8)

Single Nucleotide Polymorphisms (SNPs)

• Major source of genetic variation.

• Estimated approx. 7 million SNPs that occur with frequencies at least 5% in the human population; approx. 11 million with frequencies at least 1%.

• Can we determine the associations between these variants and diseases?

Other types of genetic variations…

International HapMap project

• Haplotype – set of variants on a chromosome that tend to inherited as a block.

• Provide a collection of SNPs spanning the genome, and serving as genetic markers.

• Study correlations (linkage disequilibrium, LD) between the SNPs.

• Provide a guide for whole genome association studies.

HapMap project

• Project was launched in Oct 2002.

• In the first phase genotyped 1.1 million SNPs in 269 individuals from four ethnic origins.

• Second phase will genotype another 4.6 million SNPs.

• Goal was to find most SNPs that occur with frequencies of at least 5% in the human population.

Statistics digression: here is an example of a commonly used correlation measure…

LD and recombination hotspots

Correlated (LD) SNPs and tag SNPs

Nature Genetics: published online Oct 30, 2005; doi:10.1038/ng1688

Haplotype diversity

Nature, v. 437 Oct 27, 2005, p.1306

LD summary

• The human genome consists of regions of low polymorphism (i.e. low sequence variation) of sizes from 10-100 kb, interspersed with regions of high polymorphism.

• This seems to be due to “recombination hotspots” in the chromosomes.

• The inheritance of chromosomal regions without recombination (haplotypes) means that certain combinations of genes are widespread across the human population.

http://www.hapmap.org/

Exercise!

• Go to www.hapmap.org, and select “Browse Project Data” (link on the left).

• In the “Landmark or Region” box enter: DTNBP1, then click “Search”.

• Select the NM_032122 link (isoform a).• Take a look at the Overview and Details.• Go down to “Tracks”, select “Analysis All on”, and then

“Update Image”.• Take a look at the LD map, phased haplotypes, and list

of tag SNPs.

Whole genome association study

• Given a sample of people, some with and some without a certain trait/phenotype (e.g. a certain disease).

• Call the two sets cases and controls.

• Investigate the genetic factors shared by the cases, but absent from the controls; i.e. find the associations between the genetic factors and the disease.

• The most straightforward way: genotype all the individuals.

• But this is far too expensive with current technology!

The HapMap data is useful for whole genome association studies…

• The collection of SNPs give us common genetic markers.

• By using tag SNPs we can reduce the number of SNPs that need to be genotyped in the study.

• It is even possible to produce SNP chips with a few hundred thousand tag SNPs that can be used for the genotyping.

• But statistical studies need to be done!