One of the major lessons from these early GWAS of depression, as with other complex traits,59,60 was that the effect of most SNPs was small in magnitude (allelic odd ratios of around 1.3 or less) and therefore considerably larger samples would be needed to identify genetic loci associated with depression. To enhance the power of psychiatric GWAS studies, the Psychiatric Genomics Consortium (PGC) was established in 2007 as an international collaborative effort to define the spectrum of risk variants across psychiatric disorders (http://www.med.unc.edu/pgc). One of the major goals of the PGC was to conduct mega-analyses for MDD in addition to schizophrenia, bipolar disorder, autism, and attention deficit hyperactivity disorder.61-63 A mega-analysis pools individual-level phenotype and genotype data from across many studies; this approach differs from a meta-analysis, where the summary statistics produced by each study are analyzed. In 2012, the PGC published the results of a GWAS mega-analysis of MDD comprising 9,240 cases and 9,519 controls across 9 primary samples, all of European ancestry.64 Although this was the largest sample to date, no SNP reached genome-wide significance. The most significant
