T15. SHARED AND DISTINCT INDIVIDUAL-SPECIFIC CORTICAL MORPHOLOGICAL PATTERNS AND THEIR ASSOCIATIONS WITH COGNITION, PSYCHOPATHOLOGY AND GENETIC RISK IN MAJOR PSYCHIATRIC DISORDERS

Behaviral and genetic similarities among schizophrenia (SCZ), bipolar disorder (BD), and major depressive disorder (MDD) suggests overlapping neural substrates. To better understand individual differences, we used an MRI-based personalized brain mapping approach to examine unique and common patterns of cortical structure across these disorders. We also explored how these brain features relate to neurocognition, clinical symptoms, and genetic risk. We derived the personalized cortical signatures, including cortical thickness (CT), grey matter volume (GV), and surface area (SA) of 327 healthy controls and 420 patients diagnosed with SCZ, BD current episode (hypo)manic or depressed, or MDD. Polygenic scores (PGSs) were calculated for SCZ, BD, and MDD, based on GWAS summary statistics from the Psychiatric Genomics Consortium. Additionally, we derived pathway-specific PGSs (pPGSs) reflecting SCZ, BD, and MDD risk through nine neurotransmitter systems: glutamate, GABA, dopamine, serotonin, opioid, histamine, norepinephrine, choline, and cannabinoid. Using a multivariate data-driven approach, partial least squares correlation (PLSC) analysis, we identified patterns of cortical alterations and their associations with IQ, symptom severity, and genetic risk. PLSC identified five latent components (LCs). A transdiagnostic CT pattern emerged across all three disorders, showing widespread cortical thinning in sensorimotor, dorsal attention, ventral attention, frontoparietal control, default mode, and limbic networks. In BD and SCZ, distinct GV and SA patterns (GV-LC1 and SA-LC1) reflected reductions in dorsal attention, limbic, and sensorimotor regions. The expression of these two patterns correlated with both IQ and GABA-related pPGSs. A depression-specific pattern (GV-LC2) in MDD and depressed BD showed increased GV in ventral attention areas. Additionally, another pattern (SA-LC2) was found most pronounced in depressed BD but least in SCZ, involving ventral attention, frontoparietal, and visual networks. The findings highlight both shared and disorder-specific alterations in cortical structure referring to different measures across SCZ, BD, and MDD, each linked to distinct symptomatic, cognitive and genetic profiles. This study suggests that individualized brain mapping can capture neuropathological signatures that cut across traditional diagnostic boundaries while also delineate disorder-specific features.