COORDINATED EPISTASIS DETECTS HETEROGENOUS PATHWAYS ACROSS PSYCHIATRIC DISORDERS AND COMORBIDITIES

Cross-disorder analyses in psychiatry often center around genetic correlation, which quantifies the average similarity of genetic effects across two disorders. For a long time, this has been the only feasible approach, as most cohorts only collect data on a single disorder. However, few studies have examined the genetic architecture of comorbidity itself or how it relates to the genetic architecture of the individual disorders involved. In this study we set out to investigate the genetic architecture of comorbidity between psychiatric disorders in the iPSYCH2015 case-cohort study. This Danish register-based study contains comorbid cases for 10 pairs of five psychiatric disorders (schizophrenia (SCZ), bipolar disorder (BPD), major depressive disorder (MDD), autism (AUT) and attention deficit hyperactivity disorder (ADHD)), making it ideal for understanding comorbidity. We develop a novel framework to model both cross-disorder genetic sharing and the genetics of comorbidity based on the concept of Coordinated Epistasis (CE). Within this framework, we can identify synergistic and antagonistic interactions of Polygenic Risk Scores (PRS) across each disorder pair. We can also identify how these interactions impact individual disorders involved and delineate established theoretical models of comorbidity. In particular, we test one model of comorbidity where genetic effects distinguish comorbid cases from cases with only one disorder, which shows synergistic PRS interactions between ADHD-AUT comorbid cases and cases of either AUT or ADHD, which replicates in both iPSYCH2015 sub-cohorts: 2012 (P = 1.3E-02) and 2015i (P = 2.9E-02). We next apply our framework to family-based genetic scores (PA-FGRS), using recorded diagnoses from an average of 20 genetic relatives from the Danish medical registry. We find synergistic PA-FGRS interactions in comorbid ADHD-AUT (P = 1.1E-05), validating our PRS results. In summary, we perform the first comprehensive study on the genetics of comorbidity by extending the CE framework using a combination of PRS and PA-FGRS, and for the first time identify coordinated polygenic interactions contributing to cross-disorder genetic sharing and comorbidity among five psychiatric disorders.