under a skewed offspring distribution among individuals in a - - PowerPoint PPT Presentation

under a skewed offspring distribution among individuals
SMART_READER_LITE
LIVE PREVIEW

under a skewed offspring distribution among individuals in a - - PowerPoint PPT Presentation

Expected time to coalescence and F ST under a skewed offspring distribution among individuals in a population Bjarki Eldon (with John Wakeley) Mathematics and Informatics in Evolution and Phylogeny June 10-12, 2008 High variance in offspring


slide-1
SLIDE 1

Expected time to coalescence and FST under a skewed offspring distribution among individuals in a population

Bjarki Eldon (with John Wakeley) Mathematics and Informatics in Evolution and Phylogeny June 10-12, 2008

slide-2
SLIDE 2

High variance in offspring distribution

  • broadcast spawning and external fertilization
  • type III survivorship curves
  • very large population sizes
  • low genetic variation
  • large number of singleton genetic variants
Surviv
  • rship
urv es

❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅

I I I I I I P er en t Life Span

#

surviv
  • rs

100

slide-3
SLIDE 3

Standard reproduction models have finite variance in offspring number Moran model of overlapping generations: a single randomly chosen individual produces one offspring Coalescent timescale: N2 2 generations

slide-4
SLIDE 4

A modified Moran model

A special case of the models of Pitman (1999) and Sagitov (1999) The number U of offspring is a random variable The probability Gn,x of an x-merger (2 ≤ x ≤ n): Gn,x =

N

  • u=2

PU(u) u x N − u n − x

  • N

n

  • PU(u) =

       1 − φN −γ/2 if u = 2 φN −γ/2 if u = ψN, 0 < ψ < 1

slide-5
SLIDE 5

Population subdivision with migration

Conservative migration between finite number D of subpopulations gives convergence to the structured coalescent if Nm is finite as N → ∞ Nγ ≡ min

  • N γ, N 2

, γ > 0 Nγ is the coalescence timescale; mγNγ < ∞ as N → ∞ mγ is rescaled migration; λγ = Iγ≥2 + φψ2Iγ≤2, φ > 0, 0 < ψ < 1 λγ is the rate of coalescence of two lines

slide-6
SLIDE 6

Time to coalescence for two lines sampled from same (T0) or different (T1) subpopulations E(T0) = D λγ < D λγ + D − 1 Nγmγ = E(T1)

λγ = Iγ≥2 + φψ2Iγ≤2, Nγ ≡ min (N γ, N 2)

slide-7
SLIDE 7

Indicators of population subdivision - FST and NST

FST defined in terms of probabilities of identity NST defined in terms of average numbers of pairwise differences FST = 1 1 + Nγmγ λγ D2 (D − 1)2 + θ/2 λγ D D − 1 , NST = 1 1 + Nγmγ λγ D D − 1 In a Wright-Fisher population: FST = 1 1 + 4Nm D2 (D − 1)2 + θ D D − 1 , NST = 1 1 + 4Nm D D − 1 λγ = Iγ≥2 + φψ2Iγ≤2, Nγ ≡ min

  • N γ, N 2
slide-8
SLIDE 8

SUMMARY (i) multiple mergers coalescent processes may better apply to some marine organisms (ii) coalescent times are shorter than in the standard coalescent (iii) patterns indicating population subdivision can be observed in DNA sequence data even if the usual migration rate Nm is very, very large