Files
Abstract
In this dissertation, two different areas of statistical genetics are explored. The first is a analysis of genotyping errors in microsatellite data and their effect on population differentiation statistics. Although, genotyping errors in microsatellite data have been explored for their effects on parentage assessment, especially with exclusion and on population size estimates of mark and recapture studies. This research is the result of need to understand the effects of inevitable errors within microsatellite data on conclusions about population differentiation. Chapter 2 illustrates the statistically significant effects that three common genotyping errors (allelic dropout, binning error, and null alleles) have on the population differentiation statistic FST. These errors however, produce no change in the overall conclusions about the differences between populations. The second is a method for improving gene mapping of complex diseases. Chapter 3 describes a process using genetical genomics methods to cluster expression level genes by their causative locus and then inferring the genotype structure of these causative loci for each individual. General association studies of a particular locus have reduced power due to individuals present whose disease is not influenced by that locus. Our inferred genotype structure is used to eliminate individuals where this is the case to increase the power of gene mapping.