UC Berkeley

Whole genome sequences allow for new, powerful inferences of evolutionary history. In addition to increasing the scale of application of traditional population genetic inferences by using them on an enormous amount of loci, genomic data also provide valuable information about the location of those sites and correlations among them, which can be leveraged for inference of evolutionary parameters. In this dissertation, I explore three types of evolutionary inferences that are thriving with the increasing availability of genomic data and new methods. In Chapter 1, I use simulations to evaluate methods for inference of ancestral recombination graphs. In Chapter 2, I apply a new method based on the pairwise sequential Markovian coalescent to infer population split times and migration rates from a pair of diploid genomes. In Chapter 3, I use population-level genomic data from the population of a rural village in Ecuador to infer signatures of selection possibly related to the beneficial effects of diet on their cardiovascular health.