For all variants with case-control genotype data available in three or more independent samples, we calculated crude ORs (Odds Ratios) and 95% C.I. values of the addiction susceptibility from the allelic distributions for each study following the published protocols [4]. Summary ORs and 95% C.I. values of the addiction susceptibility were then calculated using both the fixed-effects model and DerSimonian & Laird random-effects model [22]. We further graded the epidemiological credibility of these genetic associations according to the criteria of the Human Genome Epidemiology Network (HuGENet) [24]. Details of the grading system followed Ioannidis et al [46]. Briefly, each meta-analyzed association is graded on the basis of the amount of evidence, consistency of replication, and protection from bias, following the published protocols [4].
