Genome wide association studies (GWAS) have revealed fascinating insights into the genetics of complex diseases. These studies provide many statistical challenges but one problem that has received surprisingly little attention is the testing of associations between phenotype and genotype on the X chromosome.
In this thesis we show that there are methods that perform significantly better than those in current wide-spread use for the analysis of X chromo- some GWAS data. In particular we establish that the methods proposed by Clayton (2008) are amongst the most powerful for X chromosome analy- sis. We quantify these gains via a simulation study under a variety of genetic models and experimental designs, to compare eight existing analytical meth- ods.
Using the knowledge gained from this simulation study we apply the most powerful method to the X chromosome data from a genome wide association study of multiple sclerosis (Australia and New Zealand Multiple Sclerosis Genetics Consortium (ANZgene), 2009). Our analysis identifies 11 genetic markers that warrant further study, an improvement upon the published analysis of this data.