A Manhattan plot of our meta-analysis results is provided in Figure 1 with a quantile-quantile plot provided in Supplementary Figure 1. Linkage Disequilibrium Score (LDSC) regression 10 produced a genomic inflation factor (λGC) estimate of 1.63 with an intercept of 1.015 (0.011) prior to inflation correction, indicating that the inflation was due to polygenic signal and unlikely to be confounded by population structure. All of the 102 associated variants had the same direction of effect on depression across the three contributing studies and also within an independent replication sample of 1,306,354 individuals (Table 1; 414,055 cases and 892,299 controls). In the replication sample, 97 out of the 102 associated variants were nominally significant (P < 0.05) and 87 were significant after Bonferroni correction (α = 0.05 / 102; P < 4.90 × 10-4). Further examination of the general directionality agreement of associated variants found in the contributing studies to the meta-analysis is provided in Supplementary Table 2 and the Supplementary Note. In summary, the direction of effect of depression variants in the previous studies was consistent with the current meta-analysis.
