Combining Admixture Mapping and Association Test to Improve the Power of Genome-Wide Association Studies

• Activity Number: 588
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Jasmin Divers*+ and Nicholas M. Pajewski and David Redden and Raymond James Carroll and David B. Allison and Carl D. Langefeld
• Companies: Wake Forest University and Wake Forest University and University of Alabama at Birmingham and Texas A & M University and University of Alabama at Birmingham and Wake Forest University
• Keywords: Local admixture ; Genome-wide association studies ; joint test ; measurement error ; stistical power ; type I error rate

Abstract

Admixture mapping and genome-wide association studies (GWAS) have been applied separately with various degrees of success in gene mapping efforts. There is evidence that controlling for local ancestry in GWAS may improve the type I error rate in some cases. However, there has been very little emphasis placed on the power of combining admixture mapping with genetic association tests. We have showed in previous work how admixture mapping based on unrelated individuals can be cast as a test of linkage based on identity by descent probabilities. Here we consider various joint tests of linkage and association. We compare their performance under various admixture and population genetic models in an effort to describe cases where such an approach might useful. We also compare the magnitude of the measurement error that occurs in the estimation of the local admixture to the levels observed with the global admixture estimate, and assess their respective effect on these joint tests. We evaluate the effect of the window size selected to obtain the local admixture on the measurement error and the performance of each test.