W7 Fri GWAS for complex trait

Question 1

What do we mean by heritable

Answer 1

• Heritability is the proportion of phenotypic variation that is due to genetic variation in a population.
• Broad sense heritability: H2 = VG /VP
• Non-genetic variation is attributed to the environment: VG + VE = VP

Question 2

Way to study heritibility

Answer 2

Need to decompose VP (which is what we observe) into VE and VG
* Comparing monozygotic twins raised together and apart: VG = 0
* Comparing monozygotic to dizygotic twins raised together: VE = 0
* Correlation between twin pairs in general
* Correlation between parents and offspring

Question 3

Broad-sense and narrow-sense heritability

Answer 3

genetic component can be broken down further: VG = Va + V d + V i
* Va = additive variance
* Vd = dominant variance
* Vi = epistatic variance
* Narrow sense heritability: h^2 = Va /V p
(aka: slope of the linear regression line)
* When we talk about the genetic contribution to traits, this is often what we mean

Question 4

Method for mapping heritability

Answer 4

Linkage mapping
* Use a segregating pedigree to construct a linkage map
Candidate gene association mapping
* Test one or a handful of pre-selected loci for association with trait in cases and controls
Genome-wide association study (GWAS)
* Test cases and controls for association with many markers.

Question 5

disadvantage of linkage

Answer 5

• Linkage mapping requires extensive pedigrees to reach significance.
• If trait has complex architecture, it will be hard to identify causal variants.
• Identifies large blocks (many MB wide).
• Suited to monogenic disorders.

Question 6

Disadvantage of associating mapping

Answer 6

• Candidate gene mapping relies on a priori assumptions about trait aetiology and causality.
• Meaningless if loci misidentified.
• Much contributing variation can be missed.

Question 7

What does the success of GWAS depend on

Answer 7

• Well-differentiated cases and controls drawn from the same population
• Sufficient statistical power (sample size) to detect significant associations

Question 8

How GWAS is carried out

Answer 8

• It is too expensive to sequence the whole genome of every individual in a GWAS ($1000/person, >500,000 people in current GWAS)
• Use tag SNPs instead:
• tag SNP: a SNP that summarises variation within an LD block

Question 9

How tag SNPs work

Answer 9

• A set of tag SNPs can be used to build a haplotype, which summarises genetic diversity at a region (thanks to LD)
• If your tag SNPs don’t tag what you think they are tagging… all downstream inference will be imprecise

Question 10

Early GWAS reseaches

Answer 10

• Myocardial infarction (Ozaki et al, 2002), age-related macular degeneration (multiple groups, 2005), both identified a small number of loci with large effects on disease risk.
  -These loci could explain a substantial fraction of the h2 estimates

Question 11

Importance of power in GWAS

Answer 11

• Early GWAS were underpowered to detect most associations.
• The strength of the association between the
  trait and the genotype depends on:
• Effect size of the variant
• Penetrance of the variant
• Frequency of the variant
• Quality of the case/control separation
• (and others)
• All of these interact in complex ways:
• A rare variant of large effect will not be detected in a GWAS, unless the sample size is very large.
• Nor will a common variant of large effect but low penetrance

Question 12

Predicting traits with polygenic scores

Answer 12

• GWAS hits can be quite informative even if the underlying biology is unknown.
• Take every single SNP ever associated with a trait and count the number of risk
  alleles you see across all of them in an individual
• The combination of a PRS and clinical risk estimates could be used in the
  clinic to recommend specific interventions for each patient