The strength of the current study is that a relatively homogeneous Japanese population was used, which may have minimized the effects of differential genetic backgrounds. The major limitation of the current study is that the conservative corrections for multiple testing may have missed a large proportions of eQTL SNPs. Increasing power allows better detection of weaker and more distantly located cis-regulatory elements [23]. Greater than 82% of the significant eQTL-probe pairs identified in the current study had P<3.1×10−13, which far exceeded the predetermined significance level (P<3.1×10−12). Our findings should not be generalized to more weakly associated eQTLs since they may have different regulatory mechanisms. Another limitation is that approximately half of the samples were collected from patients with a depressive disorder. However, analyzing healthy and depressive subjects separately also resulted in achieving high correlations (FDR-corrected P<0.01) for all the 1,554 SNP-probe pairs identified in the current study. Therefore, it is unlikely that depressive disorder has a major impact on gene expression regulation of the identified eQTL SNPs. Further investigation on the influence of depressive symptoms on gene expression levels is underway using a larger sample size.
