Genome-wide association studies (GWAS) have provided insight into the genetic etiology of multiple brain disorders. However, extracting biological mechanisms from GWAS data is a challenge, largely because the majority of common risk variants reside in non-coding regions of the genome1.