A hierarchical multi-scale framework for schizophrenia: integrating symptom networks, functional circuits, and...

Schizophrenia is characterized by striking symptom heterogeneity, yet the mapping between specific clinical phenotypes and their underlying biological substrates remains elusive. To bridge this gap, we applied an integrative, multi-scale framework combining symptom network analysis, connectome-based predictive modeling (CPM), and transcriptomic mapping in a multi-center cohort of schizophrenia patients. This approach revealed a hierarchical dissociation between phenotypic topology and underlying biological mechanisms among the identified high-centrality core symptoms: Conceptual Disorganization, Unusual Thought Content, and Blunted Affect. While Conceptual Disorganization and Unusual Thought Content exhibited clinical coherence as psychosis-related features, CPM uncovered a divergence in their neural substrates. Conceptual Disorganization shared greater neurofunctional isomorphism with Blunted Affect-characterized by converging dysconnectivity within the Somatomotor Network (SMN) and subcortical circuits-whereas Unusual Thought Content displayed a distinct architecture driven by prominent Default Mode Network regulation beyond the shared sensorimotor substrate. Transcriptomic annotation further stratified these dimensions: psychosis-related networks were underpinned by synaptic regulatory genes, whereas Blunted Affect was enriched for intracellular MAPK signaling and metabolic processes. These findings delineate a hierarchical model in which distinct molecular etiologies-synaptic versus metabolic-cascade into shared systems-level failures at the "somato-cognitive interface". We conclude that while symptoms may group clinically, their treatment requires targeting separable molecular pathways that converge on common circuit bottlenecks. This framework reconciles symptom heterogeneity with overlapping biological substrates, advocating for a mechanism-based stratification of schizophrenia.