Inferring, Estimating, and Accounting for Population and Pedigree Structure in Genetic Analyses

Abstract

Genetic studies of admixed individuals with ancestry derived from multiple previously isolated populations have become more common in recent years. Statistical methodology for analyzing genetic data in the presence of complex population structure and relatedness is currently of need. In this dissertation, we thoroughly investigate the performance of existing analysis methods and explore improvements where limitations exist. Methods for accurate ancestry inference and relatedness estimation when both structures are simultaneously present are developed and then utilized to achieve improved performance of genome-wide genetic association testing for complex quantitative traits. The performance of novel methodology is assessed through the use of extensive simulation studies as well as in applications to real data sets from admixed populations.