Testing for Gene-Environment and Gene-Gene Interactions Under Monotonicity Constraints

• Activity Number: 295
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Summer S. Han*+ and Philip S. Rosenberg and Nilanjan Chatterjee
• Companies: National Cancer Institute and National Cancer Institute and National Cancer Institute
• Keywords: Gene-gene interaction ; gene-environment interaction ; isotonic regression ; constrained likelihood ratio test ; order restrictions ; genome-wide association study (GWAS)

Abstract

Recently there have been abundant reports on genetic interactions and frequently observed are reports on qualitative interactions,e.g. directions of genetic effects flip depending on exposure levels. In some contexts, however, this is believed to be biologically less plausible, and many of such findings failed to be replicated. To reduce false positives and increase power of detecting more plausible forms of interactions, we develop constrained interaction tests using bivariate isotonic regressions. Our new method imposes constraints that ensure quantitative interactions, where directions of genetic effects remain consistent regardless of exposure and vice versa. Simulations show the new method gains power over traditional methods by reducing the size of alternative parameter spaces. We show that our method is computationally feasible for GWAS by application to NCI lung cancer GWAS data.