Alcohol dependence (AD) is a significant social and economic problem. The propensity to develop AD is strongly influenced by genetics (Rodriguez et al., 1993, Heath et al., 1999, Schuckit and Smith, 1996, Prescott and Kendler, 1999, Schuckit, 2000), but few of the specific genetic factors underlying this propensity have been identified. One major problem with the identification of strong candidate genes in humans is that promising signals are frequently not replicated across studies at the level of the individual polymorphism or gene. Because variation in different genes that act together in a biological process may yield similar phenotypic outcomes, we suggest that identification of associated variation in different genes controlling the same process across studies strongly supports a role for that biological process in AD. For this reason, biological process-based set analyses may provide more reliable and replicable results than individual gene-specific analyses. Here, we assess the association of all genes known to act together to perform a common biological action with alcohol-related outcomes in an adult sample drawn from six treatment centers, the Collaborative Study on the Genetics of
