UC Berkeley

The concept of relatedness is fundamental is many areas of genetic studies, such as disease association studies, conservation genetics, and inferences about the demographic history and social structure of a population. Related individuals show signatures of shared ancestry in their genomes, which can then be analyzed to infer the relationship. In this thesis, we present statistical methods for estimating the relationship between individuals at varying time scales and the population parameters that produced such structure. In particular, we develop methods for analyzing genetic markers to estimate the pedigrees of close relatives and the mating parameters, such as the effective population size, that govern the population. Using simulations, we find that our method can infer pedigrees and the effective population size better than existing methods. We also discuss a method to infer regions of Neanderthal ancestry in human genomes, which can then be used to study the distant relationship between Neanderthals and humans. We apply the method on a sample of ancient humans to estimate the date of admixture between Neanderthals and humans.