Related Experiment Video
Updated: Aug 26, 2026
Developing Neuroimaging Phenotypes of the Default Mode Network in PTSD: Integrating the Resting State, Working Memory, and Structural Connectivity
Published on: July 1, 2014
Cross-Modal Brain Network Disruptions in Schizophrenia: Insights from Multiplex Modeling and Transcriptomic
Wei Zhao1,2, Hanrui Chen1,2, Hairong Xiao1,2
1MOE-LCSM, School of Mathematics and Statistics, Hunan Normal University, Changsha 410006, China.
Background And Hypothesis:
Conventional multimodal neuroimaging studies in schizophrenia (SCZ) typically model brain networks as independent single-layer systems, overlooking their coordinated cross-modal organization. The molecular and genetic underpinnings of these network alterations remain poorly understood.
Study Design:
We employed a multiplex network approach integrating functional, anatomical, and morphological networks in 135 individuals with SCZ and 148 healthy controls. Global network integration and local segregation were assessed using the multiplex participation coefficient (MPC) and multiplex clustering coefficient (MCC), respectively. Group differences were tested, and support vector machine (SVM) classifiers were constructed to evaluate diagnostic utility. Associations between multiplex network alterations and whole-brain gene expression were examined, followed by pathway and cell-type enrichment analyses.
Study Results:
Individuals with SCZ showed reduced cross-modal network integration, reflected by decreased MPC across cortical and subcortical regions. Alterations in local segregation, indexed by MCC, were more regionally restricted, predominantly affecting frontal and limbic systems. SVM models using MPC and MCC demonstrated robust diagnostic performance, outperforming single-layer network features. Transcriptional association analyses revealed that MPC reductions were significantly enriched for genes related to Wnt signaling and axonal and dendritic architecture, whereas MCC alterations were associated with transcriptional regulation and synaptic processes. Positively weighted genes were enriched in endothelial cells, microglia, oligodendrocytes, and vascular leptomeningeal cells, whereas negatively weighted genes were enriched in excitatory and inhibitory neurons.
Conclusions:
Multiplex network modeling uncovers cross-modal disruptions in brain organization in SCZ and links these alterations to specific molecular pathways and glial-neuronal substrates, providing multiscale insights into the disorder's neurobiology.
