Linkage disequilibrium decay and past population history in the human genome.

Thomas Mailund,Leeyoung Park

doi:10.1371/journal.pone.0046603

Abstract

The fluctuation of population size has not been well studied in the previous studies of theoretical linkage disequilibrium (LD) expectation. In this study, an improved theoretical prediction of LD decay was derived to account for the effects of changes in effective population sizes. The equation was used to estimate effective population size (Ne) assuming a constant Ne and LD at equilibrium, and these Ne estimates implied the past changes of Ne for a certain number of generations until equilibrium, which differed based on recombination rate. As the influence of recent population history on the Ne estimates is larger than old population history, recent changes in population size can be inferred more accurately than old changes. The theoretical predictions based on this improved expression showed accurate agreement with the simulated values. When applied to human genome data, the detailed recent history of human populations was obtained. The inferred past population history of each population showed good correspondence with historical studies. Specifically, four populations (three African ancestries and one Mexican ancestry) showed population growth that was significantly less than that of other populations, and two populations originated from China showed prominent exponential growth. During the examination of overall LD decay in the human genome, a selection pressure on chromosome 14, the gephyrin gene, was observed in all populations.

Highlights

Linkage disequilibrium (LD) is an important parameter in population genetics
This study investigated the expectation of LD decay involving changes of effective population size under various circumstances, and the results were applied to the human genome using HapMap phase III data to infer the past population history of each human population
The current study provides an actual expression of the expectation of LD decay depending on recombination rate and changes in effective population size

Summary

Introduction

Linkage disequilibrium (LD) is an important parameter in population genetics. The LD in the human genome has been used to determine the association between variants and traits [1], and efforts to understand selection pressures have been based largely on the LD status of populations [2–5], The theoretical basis of expectations for LD was established by the pioneering efforts of theoretical geneticists [6–14]. Continued efforts enhanced the accuracy of the expectation of r2 based on the ratio of expected values or related improvements [10–12,15–18]

Methods

Results

Conclusion