Quantitative Genetics Flashcards
What types of traits can there be?
- Continuous - e.g., height
- Meristic - e.g., counts of bristles
- Discrete - via threshold effect - e.g., genes determining liability towards a disease which may have high or low penetrance
How can we measure phenotypic variation
- Measure ‘quantitative traits’
- Contributory effects at a few separate loci can produce a quantitative effect
- Variation in environment can also produce a quantitative effect
What two main parameters can be used for statistics on traits with a normal distribution?
- Mean of distributions
- Variances of distributions - avg difference of individuals from pop mean
What is important about effects on multiple genes on the mean and variances?
They are additive:
- So mean and variances are the sum of the mean and variances for the genes
- Only if two effects act independently to determine a value
What is the Phenotypic value (P)?
- P is the trait value measured for an individual
- P = u + G+E
- P is influenced by genotypic effects (G), and environmental effects (E)
- Interactive effects (GXE) are possible - where effect of genotype depends on the environment it is in
What is total variance made up of for a trait?
Total variance (Vp) = genetic variance (Vg) + environmental variance (Ve)
How can we partition genetic variance (Vg)?
- Vd = component due to dominance
- Vi = component due to epistasis - interaction between genes
- Va = component due to simple difference between alleles (‘additive’)
- Vg = Va + Vd + Vi
What are the two types of heritability?
Broad sense heritability: H^2 = Vg/Vp
- The proportion of phenotypic variance that is due to genetic differences among individuals (all components)
Narrow sense heritability: h^2 = Va/Vp
- The proportion of phenotypic variance that is due to additive genetic variance among individuals
Only narrow sense heritability determines the response to selection - because the combinations of alleles giving rise to dominance and epistatic effects are broken up when passed between generations - but the additive effects dont change
What genetic variation does selection act on? Give an example
Selection acts on additive genetic variation - narrow sense heritability
- R = h^2xS
- R = response to selection
- S = selection differential
- e.g., beak depth in Galapagos island - draught caused increase in mean beak depth - difference between pop means - response of selection- deeper beaks allowed birds to access seeds
Why is it difficult to measure human values?
- Is difficult to separate similar environment from similar genetic background
- Heritability changes if the environmental variance changes
What are the key questions we may ask to tell us about the molecular basis of the trait?
- How many loci?
- Effect sizes of loci?
- Few loci of major effect, or many loci with small effects?
What different techniques can we use to make estimates of the number of loci involved?
- Pedigree based studies - large number of individuals and know who they are related to in many generations and we see how the phenotype covaries along the pedigree
- Genetic association studies - make link between phenotype and genotypic information for large number of individuals
Give an example of a pedigree based analysis
Hypertension - high blood pressure - causes multiple health problems
- Found strong positive correlation between inbreeeding coefficient (F) and the prevalence of hypertension
- Suggests that there are relatively few loci having a large effect and large number of loci having a small effect
- Rudan et al., 2003
How do association studies work?
- Screening of diff loci across genome - cases and controls
- The majority of loci would be segregating independently of the hypothetical gene (diff chromosomes or far apart in same chromosome)
- Loci linked to disease locus: i) segregating together with the disease in families, ii) higher linkage disequilibrium with hypothetical deleterious allele
- Association of specific SNPs/haplotypes with region carrying disease alleles
What is a GWAS?
Genome wide association study:
- Use 100s/thousands of SNPs or even whole genome sequencing…
- Complex regression models (usually likelihood analysis) e.g., positioning of QTLs between variable marker loci
- Assess significance of association with randomization (permutation) tests of marker loci
