Abstract

In the post-genomics era many questions surround the behavior of polymorphism on the genome level. However, current methodology does not allow the simulation of genomic-scale SNP maps using the coalescent. We have developed a new algorithm to simulate coalescent trees with heterogeneous recombination rates. The Argos algorithm is a novel application of sparse matrix methods and multi-precision arithmetic, and is computationally efficient, greatly increasing the number of loci that can be simulated. In order to make Argos available to the wider scientific community, this algorithm has been implemented in the SIMCORE software package, which simulates coalescent trees (gene genealogies; ancestral recombination graphs) with a user-specified genetic map. Mutations can be placed on trees according to a chosen mutation model to produce sequence data. Sequences can be analyzed in a variety of ways including the identification of haplotype blocks and QTL analysis. The software is available at http://www.stat.purdue.edu/~simonsen/simcore/.

Citations