Another major lesson from depression GWAS has been that popular candidate genes have generally not shown evidence of association. Prior to the GWAS era, meta-analyses of candidate gene studies concluded there was nominally significant evidence (at p<0.05) for six candidate genes in depression: APOE, DRD4, GNB3, MTHFR, SLC6A3, and SLC6A4.66,67 However, none of these genes nor any of the more than 100 frequently examined candidate genes have shown evidence of significant association in the published GWAS of depression to date. Replication of candidate genes in GWAS is challenging, however, as several widely studied candidate gene markers, including the serotonin transporter 5-HTTLPR variable tandem number repeat (VNTR), are not directly captured by typical GWAS platform. Some groups have developed techniques to impute or derive best-guess estimates of these genetic markers using available SNP data,68,69 though these efforts have not yet been widely adopted. However, the overwhelming evidence for many candidate genes has not been compelling.
