Two-locus Analysis for Genome-wide Association Studies

• Activity Number: 73
• Type: Contributed
• Date/Time: Sunday, August 1, 2010 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Shurong Fang*+ and Qiuying Sha
• Companies: Michigan Technological University and Michigan Technological University
• Keywords: association study ; two stages ; cluster ; joint effects ; monotonic models

Abstract

Genome-wide association (GWA) studies involve scanning markers across the whole genome to find genetic variations. For complex diseases, joint effects of genes play an important role. However, to evaluate the joint effects requires heavy computing in GWA studies. In this paper, we propose a two-stage approach to identify joint effects, especially for monotonic models. In the first stage, all markers are ranked by a single-marker test (SMT). Then we use a marker clustering algorithm to group the highly correlated markers within a certain distance and select a representative in each cluster. In the second stage, we test the two-locus joint effects of representatives. A likelihood ratio test (LRT) is used to detect joint effects. Comparing with a regular two-stage method, our method not only does reduce the computational burden, it also increases the power by reducing the number of tests.