Identifying Gene-Gene Interaction Using RNA-Sequencing Data

• Activity Number: 370
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
• Sponsor: ENAR
• Author(s): Kwang-Youn Kim*+
• Companies: Northwestern University
• Keywords: interaction ; epistasis ; GWAS ; sequencing ; association ; genetics

Abstract

A plethora of common disease genes have been identified through a combination of microarray and next-generation sequencing (NGS) technologies using GWAS. The fast pace of discovery, however, have started to plateau in the last few years partly due to oversimplifying the complexity of human diseases. A large portion of heritability is unexplained because of failure to include gene-gene interactions, or epistasis, in describing the genetic architecture of human diseases. Many studies have shown that epistasis plays a crucial role in describing the complex etiology of the human disease. We propose a statistical model that explicitly includes the epistatic terms in addition to the marginal effects for analyzing the NGS data. The model reduces the degrees of freedom compared to other methods, where the benefits are amplified when analyzing millions of SNPs. Although our model was motivated to fit the count data generated by NGS technology, it can be generalized to fit microarray data which is continuous in nature. The model also allows the gene-environment interaction terms to be included in addition to the gene-gene interactions.