In this study we report a meta-analysis of 807,553 individuals (246,363 cases and 561,190 controls) using summary statistics from three independent studies of depression, Hyde, et al. 8 (23andMe_307k), Howard, et al. 5 (UK Biobank) and Wray, et al. 9 (PGC_139k). The meta-analysis examined 8,098,588 variants and identified 102 independently segregating variants associated (P < 5 × 10-8) with depression. These 102 variants were assessed in an independent replication sample of 1,306,354 individuals (414,055 cases and 892,299 controls) and 87 variants were significant in that sample after multiple testing correction. The estimate of the genome-wide SNP-based heritability on the liability scale was 0.089 (s.e. = 0.003), indicating that the analysed depression phenotype had a significant genetic component. All of the 102 variants associated with depression had an equivalent direction of allelic effect across the three studies5, 8, 9 that contributed to the meta-analysis and the replication sample (Supplementary Table 1), suggesting that these variants represent robust associations with depression. Our gene-based analyses revealed pathways relating to neurotransmission and response to stimuli, with the central nervous system also found to be
