tissue, which is different from unique genes as genes often have GReX models available to test in multiple tissues. The most significant association signals were found in a gene dense region on chromosome 3 (Fig. 1), in line with expectation given the significant GWAS signal for AN (chromosome 3: 47 588 253–51 368 253) overlaps this cluster of significant genes. Upregulation of the gene encoding WD Repeat Domain 6 (WDR6) was the top hit in this region associated with increased odds of AN – ZTWAS = 7.44, p = 9.96 × 10−14 (GTEx cortex). This gene also surpassed Bonferroni correction in several other tissues, including the larger sample size PsychENCODE cortical samples, whole blood, nucleus accumbens, and caudate basal ganglia. The WD repeat protein family effects signal transduction (Li & Roberts, 2001), whilst WDR6 is thought to influence cell cycle arrest (Xie, Wang, & Chen, 2007). Several other proximal genes to WDR6 also survived Bonferroni correction in multiple tissues, such as, CCDC71, NCKIPSD, and MST1R. O-6-Methylguanine-DNA Methyltransferase (MGMT) was the most significant gene outside of chromosome 3 (ZTWAS = −5.33, p = 9.88 × 10−8 – caudate basal ganglia), followed by CDK11B on chromosome 1, and the long non-coding RNA
