Sex-Specific Psychosis Biotypes: Informed Data-driven Neurobiological and Genomic Markers for Early Risk Detection - Project Summary This proposal aims to address the lack of sex-specific research in psychosis, which hinders the development of sex-informed risk detection and tailored interventions that hold strong promise to improve the outcome of psychotic disorders (PDs) and reduce their substantial clinical and economical burden. However, the current diagnostic systems primarily rely on specific symptoms that may not yet be present in the prodromal stage, and do not consider sex differences. This has motivated a shift towards neurobiological and genetic markers that are known to show recognizable psychosis-related abnormalities, predisposition, and sex differences around early adolescence. First we propose to develop a novel two-mode unsupervised and supervised data-driven approach to decipher biological heterogeneity based on the brain’s functional connectivity (FC). The new approach will be applied to a large transdiagnostic B-SNIP cohort of 1200+ individuals with PD to identify subgroups in these individuals, such that these subgroups show distinct neurobiology and differential sex prevalences, denoted as sex-specific biotypes. This will provide a more comprehensive and accurate understanding of the characteristic pathophysiological brain changes in each PD subgroup, in a sex-informed manner. We will connect these characteristic pathophysiological FC changes to behavioral and cognitive measures to better understand their clinical manifestations. In addition, we will also develop a biologically-informed multivariate data fusion approach to identify genomic factors related to the pathophysiological FC changes that characterize individual PD biotypes. Particularly, we will identify both sex-differential and sex-interacting geomic-FC factors . This will improve our understanding of the sex-stratified genetic predisposition for FC alterations of individual PD biotypes. These genomic factors and PD-related FC alterations will be utilized to develop a sex-informed biotype detection toolUsing B-SNIP and our unique uncharted longitudinal dataset of 220 adolescents with first-episode psychosis. We will then test and validate this detection tool for early risk prediction using the longitudinal NAPLS-3 dataset that includes ~690 clinical high-risk individuals, and the large longitudinal ABCD cohort of 11000+ adolescents recruited at age of 10 years old, and follow-up till young adulthood. The validated biotype detection tool will be made publicly available to the research community via GitHub to facilitate early risk detection and developing tailored prevention and intervention strategies.
