Causal inference on pathophysiological mediators in psychiatry

Supported by technological advances and collaborative efforts, psychiatric genetics has provided robust genetic findings in the past decade, particularly through genome-wide association studies (GWASs). However, translating these genetic findings into biological mechanisms and new therapies has been enormously challenging because of the complexity of their interpretation. Furthermore, the heterogeneity among patients with the same diagnosis, such as schizophrenia or major depressive disorder, challenges the biological validity of existing categorical approaches in clinical nosology, which is further complicated by the pleiotropic nature of many genetic variants across multiple disorders. Therefore, in the post-GWAS era, the greatest challenge lies in integrating such enriched genetic information with functional dimensions of neurobiological measures and observable behaviors. In this integration, the causal inference from genotypes to phenotypes through intermediate biological processes is of particular importance. In this review, we aim to construct an intellectual framework in which we may obtain causal, mechanistic insights into how multifactorial etiologies—in particular, many genetic variants—affect downstream biological pathways that lead to dimensions of psychiatric relevance.