Advancing transdiagnostic data analytics using knowledge graphs

Abstract

Artificial intelligence approaches have tremendous potential to advance our understanding of biological and other processes contributing to mental illness risk. An important question is how such approaches can be tailored to support transdiagnostic investigations that are considered central for gaining deeper insight into etiological processes and psychopathology that may not align well with categorical illness delineations. Here, we present the so-called “knowledge graphs” that could be leveraged in analytic approaches to synthesize multimodal data of transdiagnostic relevance, identify important latent structures and biomarkers, and support the evaluation of existing transdiagnostic frameworks.