10: Variation in Phenotype

Phenotypic variation

Found in genetically complex traits.

On a higher level than genetic variation.

Stochastic parameters

Random parameters change the way identical genes are expressed, which can lead to different phenotypes (despite identical genotype).

Central dogma

DNA makes RNA, and RNA makes protein.

Complexity of phenotypic variation

Genes can produce alternatively spliced mRNAs, which code different proteins.

Genes in stem cells can develop into many different types of cells.

Genes can present different phenotypes (ants develop very different bodies).

Monogeneic (Mendelian) traits

Traits controlled by a single gene.

Polygenic traits

Quantitative traits; traits controlled by many different genes.

Discrete vs continuous variation

Discrete — two or more separate forms of a trait (distinct).

Continuous — an unbroken range of phenotypes (blended).

Normal (Gaussian) distribution

Many traits have a stabilizing selection; distributed in a bell-curve shaped.

Can approximate discrete traits or describe variation in multiple traits.

Quantitative genetics

Describes the inheritance of traits that depend on many genes.

Depends on sensible parameters.

Described by the additive model.

Additive model

The normal distribution is expected whenever observations are the sum of many independent and random effects.

It depends on two parameters: mean (average) and variance (spread from average).

Correlation vs covariance

Correlation — measures strength of the variables under comparison.

Covariance — measures the extent of change in one compared to another’s change.

Phenotype equation (P = G + E = A + D + I + E)

Phenotypic value = genotypic value + environmental deviation

A = sum of gene effects; D = sum of dominance effects; I = sum of epistatic effects

Environmental variance = V_E = Var(E) = Var(P)

Dominance vs epistasis

Both interactions cause deviations from the additive model (normal distribution).

Dominance — interaction between two homologous genes at a single genetic locus in diploid genomes.

Epistasis — effects where interactions between genes are not located in homologous sites, which can occur in any genomes (including haploid).

Breeding value

BV = A / 2

BV = [Offspring value — original (parental) value] / 2

Heritability equation

Var(P)  =  Var(G) + Var(E)

Var_P  =  V_G + V_E 

Heritability can be broad-sense or narrow-sense.

Broad-sense heritability

H² = V_G / V_P

If H² = 0, then only the environment contributes to phenotypic variation.

Narrow-sense heritability

h² = V_A / V_P

If h² = 0, then only genetics contribute to phenotypic variation.

Heritable variation (LBPM)

L = life history traits (lowest abundance)

B = behavior

P = physiology

M = morphology (highest abundance)

Quantitative Trait Loci (QTL)

Genetic loci that are correlated with variation in a phenotypic trait.

Can contain genes or be linked to genes.

Loci

Specific, fixed position on a chromosome where a particular gene is located.

Discovery of QTLs

Individuals from populations with different phenotypes were bred to produce F₁.

F₁ offspring were bred to produce a second generation, F₂.

Genetic markers were examined for correlations with differences in phenotypes.

Fisher’s Geomotric Model

Mutations of small effect (r << d) are as likely to improve fitness as to reduce it.

Mutations of large effect (r > 2d) must reduce fitness.

Spherical model encompasses all possible combinations, with the fittest mutations being located in the center (optimum); mutations of small effects can slowly approach optimum while mutations of large effects usually jump out of the sphere.