Efficient Forward Simulation of Fisher-Wright Populations with Stochastic Population Size and Neutral Single Step Mutations in Haplotypes

摘要

In both population genetics and forensic genetics it is important to know howhaplotypes are distributed in a population. Simulation of population dynamicshelps facilitating research on the distribution of haplotypes. In forensicgenetics, the haplotypes can for example consist of lineage markers such asshort tandem repeat loci on the Y chromosome (Y-STR). A dominating model fordescribing population dynamics is the simple, yet powerful, Fisher-Wrightmodel. We describe an efficient algorithm for exact forward simulation of exactFisher-Wright populations (and not approximative such as the coalescent model).The efficiency comes from convenient data structures by changing thetraditional view from individuals to haplotypes. The algorithm is implementedin the open-source R package 'fwsim' and is able to simulate very largepopulations. We focus on a haploid model and assume stochastic population sizewith flexible growth specification, no selection, a neutral single stepmutation process, and self-reproducing individuals. These assumptions make thealgorithm ideal for studying lineage markers such as Y-STR.