Abstract:
|
There is increasing interest in the joint analysis of multiple phenotypes in genome-wide association studies (GWASs). Multiple phenotypes might share a common genetic basis, referred to as pleiotropy in genetics. Such analysis faces several challenges: Multiple phenotypes might be correlated and are of possible high dimension. An SNP might affect only a subset of SNPs. Several approaches have been proposed to analyze multiple phenotypes, using standard multivariate analysis, dimension reduction using PCA, and variance-component based methods. We will examine analytically the factors that affect the performance of these methods, and develop omnibus robust tests that can be powerful in different scenarios. We conduct extensive simulation studies to evaluate the performance of the methods, and apply these methods to a genome-wide association study of plasma lipids levels and identify numerous novel genetic variants that conventional single-trait analysis approaches failed to discover.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.