Network connectome analysis of multi omics data identifies molecular markers of recurrence and grade progression in meningioma

Background Meningiomas are usually benign, but some behave aggressively with early recurrence. Histopathological grading alone often fails to predict outcomes. We developed a network connectome and clustering framework that integrates DNA methylation, RNA-seq, and proteomic data to identify molecular interaction patterns linked to recurrence and grade progression. Methods Using genome-wide methylation, transcriptomic, and proteomic profiles, we constructed multi-layer connectome networks representing inter-omic correlations. Nodes and edges were analyzed by centrality and clustering metrics to detect key molecular modules associated with clinical outcomes. Results Distinct network clusters differentiated recurrent and higher-grade meningiomas from indolent ones. A total of 29 methylation, 32 gene, and 33 protein features were significantly related to recurrence; 70, 61, and 56 features were linked to grade progression. Recurrent tumors showed increased inter-omic connectivity and altered hub distributions. LINC01397 emerged as a recurrent hub across omic layers, suggesting its role as a potential unified biomarker. Conclusion Our connectome-based multi-omics analysis reveals that meningioma aggressiveness is driven by coordinated molecular interactions rather than single-omic alterations. This systems-level approach provides a compact, data-driven framework for predicting recurrence and grade, supporting precision risk stratification in clinical practice.