dependence diagnoses, and to examine whether the exclusion of those cases who met criteria only for alcohol dependence altered the association. Replication was attempted in two small independent samples that contained EA and AA individuals and substance dependence phenotypes, the Study of Addiction: Genes and Environment (EA: 630 cases, 1,020 controls: AA 387 cases, 415 controls)24 and the Yale-Penn AA study (AA: 1,525 cases, 485 controls)25. Further, any GWS associations with ANYDEP in the EA sample were tested for association with alcohol intake among 452,264 individuals from the UK Biobank26 and cannabis use from a meta-analysis conducted on 184,765 individuals27. Finally, given the proposed role of reward-related neural response in the etiology of addiction28,29, we examined whether GWS loci were correlated with reward-related ventral striatum reactivity as measured with blood-oxygen-level dependent (BOLD) functional magnetic resonance imaging (fMRI) in the independent Duke Neurogenetics Study (EA n=481, AA n=118)30,31.
