Inferring Genomic Histories of Structured Populations: Lessons from the Hominids
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Cambridge
Genetics
Magdalene
Genetics
Magdalene
Abstract
Description
Demography is a major determinant of the variation of species. Understanding how it
effects the genetic composition of populations allows us to infer history from genetic data. I
contribute to this project in three ways. First, I estimate the historical trajectories of effective
population size in Pan and Pongo. These analyses are based on past rates of coalescence,
estimated within and between populations, and they use the largest known samples of whole
genomes for each species. From these, I infer new histories of migration and separation in
each genus.
Second, the interpretation of past rates of coalescence is known to depend on whether a
population is structured. I analyse a novel model of non-equilibrium island structure and
propose an approach through which it can be used to aid the interpretation of estimated
historical changes in effective population size. This is based on comparisons of the effects on
theoretical distributions of coalescent times. The approach is then applied to examples arising
in chimpanzee and human population history, in both cases suggesting effective population
size changes are more likely due to changes in census population size, rather than island
structure.
The proliferation of ancient DNA sequencing presents a rich new resource. I develop a
haplotype-based approach to help test more directly for ancestral structure. It is based on
reconstructions of the ancestral recombination graph of a sample. The method incorporates
an explicit coalescent analysis of locus ancestry under idealised demographic models. It
jointly analyses samples of ancient and modern whole genomes in which modern sequences
trace some recent part of their ancestry to populations from which the ancient sequences are
drawn. I illustrate some of the possibilities and limitations of this tentative approach and
apply it to questions about the peopling of North America.
Gates Cambridge Trust
Gates Cambridge Trust