Statistical Methods for Assessing Genetic Association in the Presence of Linkage

This thesis is concerned with the analysis of association between genetic markers and disease. We consider a scenario where it is known that a genetic region of interest has a tendency to be transmitted intact from parent to offspring. The region is said to be linked. The hope is that a mutation involved in the causal pathway of the disease is contained in the linked region, and that we can pinpoint its exact location through association analysis. We describe and assess existing methodologies, parametric and non-parametric, for the testing and estimation of association in the presence of linkage. Many genetic association studies have complex ascertainment schemes. We develop a novel score test of association in the presence of linkage for binary traits that takes ascertainment, as well as population stratification, into account.