INTEGRATING GENETIC VARIANCE AND BRAIN ACTIVITY DATA INTO A MULTI-MODAL LANDSCAPE OF PSYCHIATRIC DISORDERS

Psychiatric disorders often present substantial levels of comorbidity and shared symptoms. This can be partially explained through shared genetic factors and affected pathophysiological aspects, including impairments in structural and functional connectivity. These shared and distinct pathophysiological features are typically explored within specific modalities such as psychiatric genetics or brain imaging. However, such approaches typically result in unimodal perspectives, whereas efforts to combine them are rather challenging and uncommon. In this study, we address this gap by integrating multimodal brain data to gain more holistic insights into how genetic variation influences brain activity in health and disease. For comprehensive integration and mining of this multi-modal genetic and brain data space, we adopted a novel approach based on a Genetic Algorithm for Generalized Biclustering (GABi). This identifies hubs where genetic variation contributes to increased susceptibility or resilience to psychiatric disorders. To pinpoint genes associated with task-specific brain activity maps and other functional networks spanning various relevant domains, we explore genes associated with selected psychiatric disorders and latent genetic factors that exhibit high spatial correlations with these networks. Subsequent gene/brain activity network analyses reveal biological characteristics that either strengthened network flexibility and resilience or increased vulnerability towards psychiatric disorders. This workflow exposes pathophysiological networks that could not be identified through genetic or connectivity analyses alone. This highlights the potential of methods to integrate and mine data across different modalities, which in the future may aid in refining diagnostic boundaries and treatment options.