Complex Population Models with Coalescent Simulations

Simulated single nucleotide polymorphism data play an important role in human ge-netics. They provide a mechanism for analyzing genetic patterns on a genomic scale, examining theories of human evolution, inferring properties of recombination, and comparing efficiencies of medical genetic study designs. However, the simple popu-lation models used in most experiments pose a critical obstacle to drawing reliable inferences from these simulations. This thesis introduces a powerful and flexible tool, CoSi, that generates haplo-type data based on the coalescent process. CoSi includes key features for accurate simulation of empirical data: support for non-uniform recombination and the ability to model multiple populations and population events. Here we use CoSi to examine models of human recombination and demonstrate its capabilities by matching diverse empirical data