ALLELIC HETEROGENEITY ACROSS PSYCHOTIC DISORDERS AND RELATED PHENOTYPES

Background Major mental illnesses have been shown to overlap at the clinical and genetic levels. The genetic overlaps have been so far explored at the single genetic variants level, but very few studies have explored how independent variants within a locus could contribute to the genetic overlaps. In our study, we intend to increase the information captured from GWASs by focusing on allelic heterogeneity, i.e. the contribution of several independent markers within one genetic locus, within a trait and across related traits. Methods Using summary statistics from GWASs of traits related to mental illnesses: psychotic disorders, cognitive traits and brain volumes, we first selected independent genomic regions associated in each traits after conditional regression (Yang et al. [1] ). All the genetic variants in LD with the associated signal were included in the genomic regions. We then first explored the overlaps in the regions within traits and across traits. We also scored each genomic region in each of the traits, using the Brown score for each bin, and explored the overlap in the significant regions. Results We observed allelic heterogeneity within and across traits. 147 genomic regions were associated with independent markers (not in LD) across several traits. We have established a map of genetic overlaps for these clusters across psychiatric disorders and relevant phenotypes (brain volumes, cognitive and personality traits). The strongest overlaps were observed in pairs: schizophrenia - educational attainment and schizophrenia – bipolar disorder. We have established a pipeline for identification of allelic heterogeneity across different phenotypes. Several of the GWAS included were too limited in power to provide significant hits yet, and will need bigger samples to yield more significant results. Discussion We identify allelic heterogeneity across traits, demonstrating that some genetic regions harbor independent associations with related phenotypes. Our approach is complementary to studies that explore genetic overlap at the single marker level. This improves our understanding of the impact of genetic factors in main psychotic disorders and related phenotypes, and could help to direct functional studies later.