Gene-Environment Interactions via Algebraic Statistics, Markov Bases, and MCMC

• Activity Number: 468
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Marepalli Rao*+ and Subramanian Venkatesan and Jenny Chen
• Companies: University of Cincinnati and University of Cincinnati and University of Cincinnati
• Keywords: Single nucleotide polymorphism ; Interaction ; Cochran-Mantel-Haenszel test ; Algebraic Statistics ; Markov bases ; MCMC

Abstract

A critical step in genetic analysis is to understand how environment modulates association between a single nucleotide polymorphism and a phenotype of interest. The investigation falls under the realm of gene-environment interactions. Traditionally, Cochran-Mantel-Haenszel test is used to test interactions. For an applicability of this test certain conditions have to be met. If the conditions are not met, we have to explore the possibility of developing an exact test. In this presentation, we will introduce an exact test a la Fisher. This test leads to monumnetal computational problems. We will explain how algebraic statistics, markov bases, and MCMC will sooth computational complexities. The results are applied to a data set on metabolic sgenerated syndrome from the HVAR island in the Adriatic sea off the coast of Croatia.