SNPs showed the same direction of effect. In the SAGE GWAS the strongest signal, rs67666182, along with highly correlated SNPs (R2 > 0.8) near FABP5 showed no replication in COGA, while SNPs in LMO1, and near TLR1, TLR10 and A2BP1 showed nominal replication. Because a large excess of SNPs that were nominally significant (p<10−4) in either COGA or SAGE showed the same direction of effect in the other study, this suggests that within these SNPs there is a reproducible signal of variants that affect maxdrinks.
