GWASs represent the most recent paradigm change for gene discovery. The allure of GWASs is that they allow for interrogation of hundreds of thousands to over 1 million SNPs across the human genome at relatively modest cost. Thus, GWASs can potentially lead to the identification of variants of modest effect size that may not be recognized in candidate gene studies for their biological significance in alcoholism. The principal disadvantage is the fairly severe multiple-testing burden (i.e., the need to account for the possibility that when 1 million tests are conducted, some positive results will be obtained due to chance alone) imposed by GWASs, which results in the requirement for statistical significance to be denoted by P values of 5 × 10−8 and lower. This represents a very high significance level for validation; consequently, many SNPs that truly are associated with alcoholism risk may not yet have been recognized because they have not yet surpassed this threshold.
