Gene-gene interaction analysis of high-dimensional genomic data

Abstract

With the development of high-throughput genotyping and sequencing technology, there are growing evidences of association with genetic variants and common complex traits. In spite of thousands of genetic variants discovered, such genetic markers have been shown to explain only a very small proportion of the underlying genetic variance of complex traits. Gene-gene interaction (GGI) analysis and rare variant analysis is expected to unveil a large portion of unexplained heritability of complex traits. In GGI, there are several practical issues. First, in order to conduct GGI analysis with high-dimensional genomic data, GGI methods requires the efficient computation and high accuracy. Second, it is hard to detect GGI for rare variants due to its sparsity. Third, analysing GGI using genome-wide scale suffers from a computational burden as exploring a huge search space. It requires much greater number of tests to find optimal GGI. For k variants, we have k(k-1)/2 combinations for two-order interactions, and nCk combinations for n-order interactions. The number of possible interaction models increase exponentially as the interaction order increases or the number of variant increases. Forth, though the biological interpretation of GGI is important, it is hard to interpret GGI due to its complex manner. In order to overcome these four main issues in GGI analysis with high-dimensional genomic data, the four novel methods are proposed. First, to provide efficient GGI method, we propose IGENT, Information theory-based GEnome-wide gene-gene iNTeraction method. IGENT is an efficient algorithm for identifying genome-wide GGI and gene-environment interaction (GEI). For detecting significant GGIs in genome-wide scale, it is important to reduce computational burden significantly. IGENT uses information gain (IG) and evaluates its significance without resampling. Through our simulation studies, the power of the IGENT is shown to be better than or equivalent to that of that of BOOST. The proposed method successfully detected GGI for bipolar disorder in the Wellcome Trust Case Control Consortium (WTCCC) and age-related macular degeneration (AMD). Second, for GGI analysis of rare variants, we propose a new gene-gene interaction method in the framework of the multifactor dimensionality reduction (MDR) analysis. The proposed method consists of two steps. The first step is to collapse the rare variants in a specific region such as gene. The second step is to perform MDR analysis for the collapsed rare variants. The proposed method is applied in whole exome sequencing data of Korean population to identify causal gene-gene interaction for rare variants for type 2 diabetes (T2D). Third, to increase computational performance for GGI in genome-wide scale, we developed CUDA (Compute Unified Device Architecture) based genome-wide association MDR (cuGWAM) software using efficient hardware accelerators. cuGWAM has better performance than CPU-based MDR methods and other GPU-based methods through our simulation studies. Fourth, to efficiently provide the statistical interpretation and biological evidences of gene-gene interactions, we developed the VisEpis, a tool for visualizing of gene-gene interactions in genetic association analysis and mapping of epistatic interaction to the biological evidence from public interaction databases. Using interaction network and circular plot, the VisEpis provides to explore the interaction network integrated with biological evidences in epigenetic regulation, splicing, transcription, translation and post-translation level. To aid statistical interaction in genotype level, the VisEpis provides checkerboard, pairwise checkerboard, forest, funnel and ring chart.