Notably, in the PGC SCZ2 paper detailing the 108 SCZ-associated loci the authors attempted to ask if any eSNPs from brain eQTL databases existing at the time were credibly associated with schizophrenia. Specifically, they tested if the most significant eQTL SNP for any gene is among those SNPs 99% most likely to be credibly causal for SCZ at any locus (assuming only a single causal SNP per locus). This process led to 3 genes based on brain eQTL: 1, TINAGL1, and LIG1. In our CMC eQTL data, there is only overlap between the SCZ association and eQTLs for MLH1, with up-regulation of expression predicted to be associated with the genetic risk variation; however, this overlap is below the genome-wide threshold, Sherlock P = 6 × 10−5, Bonferroni corrected P = 0.69.
