The scale of data needed for robust findings in this area is much larger than the number of genotyped individuals in COGA alone, and COGA has contributed its genetic and phenotypic data (or summary statistics) to many larger‐scale GWAS meta‐analysis efforts, including the Psychiatric Genomics Consortium (whose Substance Use Disorders working group is led by 2 COGA members, Agrawal and Edenberg), the Externalizing Consortium (led by another COGA member, Dick), the Genetics of Personality Consortium, and the ENIGMA‐EEG consortium 98 (see, 1. Overview). However, most large scale, consortium‐based GWAS must rely upon a narrow set of phenotypes that are common across all cohorts. COGA's deep phenotyping has enabled us to delve into the findings emerging from these consortia to probe mechanisms. For instance, COGA data showed that rs1229984, the missense variant in the alcohol dehydrogenase 1B gene (ADH1B) that has been robustly associated with AUD and with alcohol consumption, was not uniformly associated with all AUD criteria but rather was driven by criteria indexing tolerance and desire to quit or cut back. 61 Although the frequency of this variant is
