With over 20,000 research participants, ours is by far the largest genetic study of AUDIT. By using a self-report measure of alcohol misuse, as opposed to recruiting a clinically-diagnosed population, we were able to rapidly and inexpensively ascertain a large number of participants. We identified rs141973904 (Supplementary Fig. 2) in the ADH cluster on chromosome 4q23, which has been previously associated with AUD (Edenberg et al. 2010; Frank et al. 2012; Gelernter et al. 2014; Treutlein et al. 2009). The same SNP has recently been associated (P = 1.22 × 10−15) with alcohol consumption using 53,089 males of European ancestry (Clarke et al. 2017). Furthermore, the most associated signal, rs182344113, which resides near the KCNJ9 (GIRK3) gene, was unknown, and is consistent with mouse studies of the homologous gene. The signal at rs182344113 was not significant and it will have to be replicated. We also identified a number of genetic correlations that have behavioral precedents, such as lifetime tobacco use, and several others that were unexpected, including lower BMI and obesity rates, and higher education. We found that AUDIT was
