PgmNr P2024: Two locus allele frequency statistics with demography and selection using a diffusion approach.

Authors:
Aaron Ragsdale; Ryan Gutenkunst


Institutes
University of Arizona, Tucson, AZ, USA.


Abstract:

Patterns of polymorphism within and between populations are determined by the demographic and evolutionary history of those populations. By measuring the present genetic diversity and patterns of linkage disequilibrium within a population, we may infer its demographic history and patterns of selection. The allele frequency spectrum (AFS) is an informative summary statistic of observed genetic variation within a sample, and has been widely used in demographic inferences for single and multiple populations, as well as to infer the effects of selection on that observed variation. AFS-based methods assume independence between sites, and the AFS does not describe levels of linkage disequilibrium between nearby sites. However, frequencies of neighboring alleles are correlated, and selection on linked sites affect patterns of observed variation, so observed levels of linkage disequilibrium are informative about the demographic and evolutionary history of the population. We present a novel numerical approach to the forward-in-time diffusion model for the joint distribution of two-locus allele frequencies and linkage disequilibrium, which can model the simultaneous effects of demography and selection on one or both loci. This allows for improved inferences on single-population demography, and describing the expected patterns of allele frequencies subject to natural selection at linked sites, including the effects of background selection and hitchhiking events.