al., 2006). In recent years, genome-wide association studies (GWASs; Treutlein et al., 2009; Bierut et al., 2010; Edenberg et al., 2010; Heath et al., 2011) also reported risk loci for alcohol dependence (summarized previously in Zuo et al., 2012). In particular, using the datasets of the SAGE (the Study of Addiction: Genetics and Environment), COGA (the Collaborative Study on the Genetics of Alcoholism) and OZ-ALC (the Australian twin-family study of alcohol use disorder) separately, several GWASs have detected some risk loci for alcohol dependence and alcohol consumption in subjects of European and African descents (Bierut et al., 2010; Edenberg et al., 2010; Chen et al., 2011; Heath et al., 2011; Wang et al., 2011) (summarized previously in Zuo et al., 2012). Different from these previous studies that reported the top-ranked risk SNPs for alcohol dependence, we reanalyzed the SAGE, COGA and OZ-ALC data, using some new analytic strategies with the goal of identifying replicable risk genes for alcohol dependence. Using the European-Americans as the discovery sample and the African-Americans as the validation sample, we identified KIAA0040 as risk gene for alcohol dependence (Zuo et al., 2012). Using the African-Americans as the discovery sample and the European-Americans as the validation sample,
