Joint Analysis of SNP and Gene Expression Data in Genome-Wide Association Studies

• Activity Number: 120
• Type: Topic Contributed
• Date/Time: Monday, July 30, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Yen-Tsung Huang*+ and Xihong Lin and Tyler VanderWeele
• Companies: Harvard School of Public Health and Harvard University and Harvard School of Public Health
• Keywords: Causal Inference ; Data Integration ; Mediation Analysis ; Multiple Marker Analysis ; Omnibus Test ; SNP-set Analysis

Abstract

Genome-wide association studies (GWAS) have been a common practice in assessing the association between single nucleotide polymorphisms (SNPs) and disease phenotype. Here we propose to exploit information on gene expression to test the link between SNPs and disease. The relations among SNPs, gene expression and disease are modeled in the framework of causal mediation modeling. Using counterfactual approach, the direct and indirect effects of SNPs on disease can be derived. Furthermore, we propose a variance component test to investigate whether there exists an overall effect of SNPs on disease by borrowing gene expression information. The test statistic under the null follows a mixture of chi-squared distributions, which can be approximated with the scaled chi-squared distribution or with perturbation procedure. The relative performance of the tests for different disease models depends on the underlying true model. To accommodate different scenarios, we also construct an omnibus test. In both simulation studies and the MRC-A asthma data, our proposed test performs well and the omnibus test can almost reach the optimal power, in which the disease model is correctly specified.