Gene-based analyses reveal novel genetic overlap and allelic heterogeneity across five major psychiatric disorders

Studies using genome-wide association (GWA) single nucleotide polymorphism (SNP) level data have indicated genetic overlap across the five major disorders in the Psychiatric Genomics Consortium (PGC): attention deficit hyperactivity disorder (ADHD), autism spectrum disorder (ASD), bipolar disorder (BPD), major depressive disorder (MDD), and schizophrenia (SCZ). However, such SNP-level analyses reveal little about the underlying biology and are reliant on correlated SNP effects across disorders. In contrast to SNPs, genes are more closely related to biology and gene-based tests can incorporate allelic heterogeneity. This study aimed to extend genetic overlap analysis across the five disorders from SNP level to gene level using a novel gene-based approach. Gene-based tests for association were performed using PGC GWA summary results for the five disorders in samples including 33,332 cases and 27,888 controls of European ancestry. After accounting for non-independence of gene-based test results, we determined whether the proportion of genes with association across multiple disorders was more than expected by chance. Similar to previous SNP-level analyses, we observed significant pairwise genetic overlap between ASD, BPD, MDD and SCZ. However, our approach also produced evidence for genetic overlap between ADHD and ASD, ADHD and BPD, and ADHD and MDD. Combining gene-based association results across disorders, 36 genes produced genome-wide significant P values (<3.2 × 10^-6). Pathway analysis of genes with P values <1.0 × 10^-3 highlighted magnesium ion binding and transport, as well as signal peptide processing, and provide insight into the biological mechanisms underlying these major psychiatric disorders.