Several pieces of evidence supported our conclusion. First, within 10Mb range surrounding this genome-wide significant risk SNP, all association signals for alcohol and nicotine co-dependence with p<10−4 were concentrated within a narrow region surrounding this SNP. This region was completely located between HTR1A and IPO11. It is, thus, highly likely that the putative causal variant for alcohol and nicotine co-dependence was located within this region. Second, many risk SNPs in this region had significant cis-acting regulatory effects on mRNA expression both in the PBMC and in the brain, increasing the possibility that the IPO11-HTR1A region plays a direct functional role in the disorder. Third, many associations discovered in European-Americans were replicated in European-Australians, and meta-analysis showed that four SNPs reached the genome-wide significance level. Some associations in European-Americans were also replicated in African-Americans. Finally, this region was specific for alcohol and nicotine co-dependence, not for any other non-alcoholism neuropsychiatric disorder examined. This region has been suggestively associated with alcohol dependence (75.6% nicotine dependence) in the same dataset before (p=2.3×10−6 by Bierut et al. 2010; p=2.8×10−7 by Zuo et al. 2011),
