In marked contrast to our OPRD1 findings, our results do not provide support for association with heroin dependence involving OPRM1 and OPRK1 SNPs. The results of our study and multiple meta-analyses do not support an association of rs1799971 with heroin dependence. We also failed to replicate the association of two OPRM1 and one OPRK1 SNP reported by Levran et al (2008) although our comparison with neighborhood controls found a nominal association with rs3778151. Considering the additional power provided by our considerably larger sample and the lack of consistently replicated findings for polymorphisms in these genes, our results suggest that it may be prudent to focus attention preferentially in future association studies on OPRD1 over the other two opioid receptor genes. However, it remains possible that significant heroin dependence risk associated with OPRM1 and OPRK1 SNPs was not detected in the current study due to small effect size, low MAF in our sample, greater variability across populations of differing ethnicity for some of these SNPs (e.g., rs6473797, rs3778151), or our failure to genotype the specific risk-associated SNPs. Given the well-documented involvement
