The genes identified using approaches (1) or (2) each overlap with genes identified in several other GWA datasets for substance dependence, based on Monte Carlo simulations ([36] and Materials and Methods). There are significant overlaps with data from: a) 600k GWA studies of a subset of these polysubstance abusers (p<0.001) [16], b) 500–600k GWA data from studies of methamphetamine dependent Japanese samples [17] (p = 0.04), c) 100k GWA data from studies of alcohol dependence in European-Americans [15] (p = 0.0003) and d) 38k data from comparisons between more frequently nicotine dependent vs less frequently nicotine dependent groups of smokers [37], [38] (p = 0.02). We also identify substantial overlap with individually genotyped data from dbGAP samples of cocaine dependent vs control individuals (Table 2). This overlap is even more impressive when we compare the genomic regions identified by clustered, nominally significant results from samples of the same racial/ethnic group (Drgon et al, in preparation).
