High-Order Interactions in Rheumatoid Arthritis Detected by Bayesian Method using Genome-Wide Association Studies Data

Abstract

Problem statement: In order to reveal the missing genetic component of Rheumatoid Arthritis (RA) susceptibility, we carried out a genome-wide high-order epistatic interaction study for RA. Approach: A powerful Bayesian strategy was applied to analyze the data of Genome-Wide Association Studies (GWAS) from the Welcome Trust Case Control Consortium (WTCCC), where 319 high-order interactions were found across the whole genome and many of which were validated by the GWAS data from the North American Rheumatoid Arthritis Consortium (NARAC). Results: This is the first study intensively searching for high-order epistatic interactions genome-widely for RA. Conclusion: Our results suggest that high-order interactions might explain a big proportion of missing genetic component of RA. In the meanwhile, synapse, calcium ion binding and membrane part likely have interactive associations with RA. This finding implies that not only autoimmune system but also nervous system can play an important role in RA.