Population structure and migration
Oct 01, 2015
Population structure and migration
Population structureRecall the island example in which isolated populations fixed different alleles by genetic driftThe resulting set of subpopulations represent a structured population in which allele frequencies vary across subpopulationsHow does structure arise, what reduces structure and how can we describe structure?
Origins of population structureGenetic drift - isolated populations drift independently to different allele frequenciesSelection in which different subpopulations experience selection favouring different alleles
Drift
Drift
Measuring population structureWe saw F when we considered inbreedingWright actually created a series of F statistics to describe various deviations from Hardy-Weinberg expectationsFor a set of subpopulations, compare the level of heterozygosity expected if they were a single random mating population (HT) to what we actually observe (HS)
Measuring population structure
Measuring population structureFST usually described as a reduction in heterozygosity
Much more intuitive to consider FST as a measure of differences between populations
Measuring population structure
What reduces FST?Migration (gene flow) moves alleles from one population into anotherTherefore migration must reduce the variance in allele frequencies among populationsMany models of migration
Continent-Island modelLarge populations sends migrants to a small population at a rate mone way migration
Island approaches continent allele frequency at a rate determined by m
Continent-Island model
Island modelSeries of similar sized populations exchange migrants at a rate m
Populations all approach the average allele frequency at a rate determined by m
Stepping-stone modelSeries of similar sized populations exchange migrants at a rate mBut migrants only move between adjacent populations
Populations all approach the average allele frequency at a rate determined by m
But adjacent populations more similar and allele frequencies can exist in a gradient
Isolation by distance
Migration limits divergenceDrift continuously increases population differences in allele frequencies, migration reduces differencesprocesses will reach a migration-drift balance:
where Nm is the total number of migrants
Migration limits divergenceMigration is very effective at removing population differences
Population structure and migration can create linkage disequilibriumModel selection favouring a dominance allele (where s is the selection coefficient)
Population may be linked by migration Called a structured population Different population will have different allelic frequencies
**Intermediate step is more what we see in nature Populations arent completely separate
Variance is as high as it can be in the bottom example
Could be same allele frequency as the top (bottom) BUT there is lower heterozgosity due to subdivision, than you had to begin with
Average heterozygosity in bottom is zero Relative to what is expected 2pq (hardy-weinberg) Would expect about 50% heterozgosity with the knowledge of allele frequencies
*Single population chopped in half and then allowing the two sides to drift Cutting off the migration
*F probability two alleles are identical be descent Gain homozygosity and lose heterozygosity BUT dont change allele frequencies If everyone mates there is no change in allele frequency Same allelic variation as pedigree
With DRIFT increase of heterozygosity if because allele frequencies are changingSmall population with increase in autozygosity lisung allele variation *F proportional reduction in heterozygosity (compared to that expected in a random mating population) Fis measure of inbreeding Within set of randomly mating individuals subpopulation this is the heterozygosity (Hexp)
Just written as F this is Fis
-----
Two subpopulations Overall level of heterozygosity observed (average of these two subpopulations) Equal allele frequencies and randomly mating two allelic variation should be the same Ht what would I expect in a set of populations, given the allelic frequencies (if they were subdivided)Hs seen, as they are subdivided
*Fst= Variance in allele frequencies across populations / normalisation
*Fst is 0 on the right Fst is 1on the left biggest difference you can have in allele frequencies
Zero to one when it is as biggest as you can get half fixed for one allele and other half fixed for the other
*Drift through time increases Fst Natural selectiongenetic drift
Migration is the OPPOSITE counteracting the process of drift as it is moving the alleles in the population This reduces the variance in allele frequencies among populations
*