TEAM: efficient two-locus epistasis tests in human genome-wide association study

Abstract

As a promising tool for identifying genetic markers underlying phenotypic differences, genome-wide association study (GWAS) has been extensively investigated in recent years. In GWAS, detecting epistasis (or gene–gene interaction) is preferable over single locus study since many diseases are known to be complex traits. A brute force search is infeasible for epistasis detection in the genome-wide scale because of the intensive computational burden. Existing epistasis detection algorithms are designed for dataset consisting of homozygous markers and small sample size. In human study, however, the genotype may be heterozygous, and number of individuals can be up to thousands. Thus, existing methods are not readily applicable to human datasets. In this article, we propose an efficient algorithm, TEAM, which significantly speeds up epistasis detection for human GWAS. Our algorithm is exhaustive, i.e. it does not ignore any epistatic interaction. Utilizing the minimum spanning tree structure, the algorithm incrementally updates the contingency tables for epistatic tests without scanning all individuals. Our algorithm has broader applicability and is more efficient than existing methods for large sample study. It supports any statistical test that is based on contingency tables, and enables both family-wise error rate and false discovery rate controlling. Extensive experiments show that our algorithm only needs to examine a small portion of the individuals to update the contingency tables, and it achieves at least an order of magnitude speed up over the brute force approach.Contact: xiang@cs.unc.edu

Flexible and robust co-regularized multi-domain graph clustering

“…The genetic interaction network is con- Figure 8: Protein-protein interaction network, gene coexpression network, genetic interaction network and crossdomain relationships structed using a large-scale Hypertension genetic data [10], which contains 490032 genetic markers across 4890 (1952 disease and 2938 healthy) samples. We use 1 million top-ranked genetic markerpairs to construct the network and the test statistics are used as the weights on the edges between markers [36]. The constructed heterogenous networks are shown in Fig.…”

Ultrafast genome-wide scan for SNP–SNP interactions in common complex disease

“…Among the methods that directly test for statistical epistasis, we report TEAM (Zhang et al 2010a) and FastEpistasis (Schüpbach et al 2010). The authors of FastCHI (Zhang et al 2009), FastANOVA (Zhang et al 2008), COE (Zhang et al 2010b) and TEAM presented a review in which TEAM was reported as the most appropriate for handling human data sets, and was therefore chosen to represent the family of methods.…”

Gene–Gene Interactions Detection Using a Two-stage Model

“…We then compute the difference in the average trait value between the two groups to determine whether the pair of SNPs is significantly associated with the trait. Finding an association between a trait and a pair of SNPs is called the ''pairwise association test,'' and recently, several different methods have been proposed for pairwise association tests (Evans et al, 2006;Zhang et al, 2010;Prabhu and Pe'er, 2012;Yang et al, 2009;Millstein et al, 2006;Ljungberg et al, 2004).…”

“…Most of these methods are only applicable to case/control data. These methods include TEAM (Zhang et al, 2010), which uses a dynamic programming approach to identify significant pairs, SIXPAC (Prabhu and Pe'er, 2012), which utilizes a novel randomization technique, BOOST (Wan et al, 2010), which utilizes a Boolean representation of data and a screening stage to filter out most nonsignificant SNP interactions, and RAPID (Brinza et al, 2010), which utilizes geometric properties of the v 2 statistic to identify significant pairs. However, none of these methods are applicable to quantitative traits.…”