The Yale-Penn genetic studies of opioid, cocaine, and alcohol dependence19,23,24 were recruited at 5 U.S. sites. All participants gave written informed consent as approved by each site’s IRB. We addressed design and assessment differences (Supplementary Tables 2 and 3, and Supplementary Methods) that prevented defining phenotypes identical to those in CATS by using an extreme discordant approach. We operationalized the ODE group as opioid dependent individuals whose opioid use had been daily or near daily, included heroin, and injection at least 100 times lifetime and the OUIP group as individuals reporting heroin use who met no lifetime DSM-IV opioid dependence criteria. We limited inclusion to European ancestry (EA) participants to examine confirmation in a sample of comparable ethnicity, retaining 643 ODE and 157 OUIP individuals for analysis.
