We further examined the entire IPO11-HTR1A region (1.5Mb) in multiple populations, and detected many association and functional signals (Tables 1, 2, 3 and S2). In the European-American discovery cohort, among 2,726 SNPs including 261 originally-genotyped SNPs and 2,465 imputed SNPs, 381 SNPs were nominally associated with alcohol and nicotine co-dependence (p<0.05) (Table 4); 57 SNPs were significantly associated with alcohol and nicotine co-dependence after region- and cohort-wide correction (α=3.6×10−6). As mentioned, one of the SNP showed evidence for genome-wide significance (rs7445832; p=6.2×10−9). All risk alleles of these markers were minor alleles (f<0.5). If conditional on the most significant SNP (i.e., rs7445832), all of the associations with other SNPs became less significant (all p>10−4; Figure 2C). In the Australian replication cohort, 100 SNPs were nominally associated with alcohol and nicotine co-dependence (0.001 ≤ p ≤ 0.049; data not shown). Thirty-four associations in the discovery cohort (6.2×10−9 ≤ p ≤ 0.049) were replicated in the Australian replication cohort (0.001 ≤ p ≤ 0.049) (Table 2 and Figure 2D), with the same directions of gene effects in both cohorts. Meta-analysis showed that all
