Introduction to statistical methods in genome-wide association studies

Abstract

Introduction After the completion of the Human Genome Project (Lander et al ., 2001; Venter et al ., 2001) and initiation of the International HapMap Project (Sachidanandam et al ., 2001), genome-wide association studies (GWAS) were designed to survey the role of common genetic variations in complex human diseases. It was expected that GWAS would have the advantage of not relying on prior knowledge of biological pathways compared with “candidate gene” studies (Tabor et al ., 2002; Wang et al ., 2005), because it assays a dense set of single-nucleotide polymorphisms (SNPs) across the whole genome. This advantage allows GWAS to overcome the bias of “candidate gene” studies due to incomplete prior knowledge. It was also expected that GWAS would have higher power and finer resolution to identify genetic variants of modest effects compared to family-based linkage studies (Risch & Merikangas, 1996). The success of identifying genes for age-related macular degeneration (AMD) under the GWAS paradigm (Klein et al ., 2005) convinced the genetics community on the efficiency and feasibility of the GWAS approach to identify unknown disease-associated variants. This study used a commercial genotyping array and assayed about 100,000 SNPs throughout the human genome. It identified the association of complement factor H (CFH) with AMD. The success of finding a common risk allele with an odds ratio (OR) of 4.6 in a small sample set of 96 cases and 50 controls has generated considerable excitement in the genetics community. The p -value of the strongest SNP association surpassed the genome-wide significance threshold after the Bonferroni correction. More importantly, this finding was replicated in the following-up studies (Donoso et al ., 2010). Undoubtedly, this encouraging finding raised the confidence among researchers to detect genetic variants that underlie various complex diseases through GWAS. In 2007, the Wellcome Trust Case Control Consortium (WTCCC) published the results of seven GWAS, including Bipolar Disorder, Coronary Artery Disease, Crohn's Disease, Hypertension, Rheumatoid Arthritis, Type 1 Diabetes, and Type 2 Diabetes (The Wellcome Trust Case Control Consortium, 2007). The WTCCC study is considered the starting point of large-scale GWAS (Visscher et al. , 2012). Since then, an increasing number of GWAS have been conducted and over 10,000 loci have been reported to be significantly associated with at least one complex trait (see the web resource of GWAS catalog (Hindorff et al., 2009), http://www.genome.gov/gwastudies/).