Multi-omics integration reveals m6A epitranscriptomic regulators as central drivers of a prognostic immune-metabolic axis in glioblastoma.

Glioblastoma (GBM) is the most aggressive primary brain malignancy, with a median survival of approximately 15 months despite current standard-of-care treatment. N6-methyladenosine (m6A) RNA modification regulates gene expression, but its relationship with miRNA-mediated regulation, metabolism, and prognosis in GBM remains incompletely understood. We hypothesized that integrating m6A regulators, miRNA-mRNA interactions, and clinical data could identify molecular relationships associated with GBM prognosis. We analyzed RNA-seq, miRNA expression, and clinical data from 160 matched TCGA-GBM patients using survival-based differential expression, Pearson correlation, network analysis, and Cytoscape visualization. A ten-gene Cox regression prognostic signature was developed and evaluated using Kaplan-Meier and time-dependent ROC analyses and externally validated using GSE83300. Functional enrichment and Hallmark Gene Set Enrichment Analysis (GSEA) were performed to characterize associated biological pathways. We identified 173 survival-associated mRNAs and 87 miRNAs, with 682 significant m6A-mRNA associations and 151 negatively correlated miRNA-mRNA pairs forming a network of 833 edges across 189 nodes. ALKBH5, WTAP, and YTHDF1 emerged as highly connected transcriptomic hubs. The ten-gene signature was significantly associated with overall survival (HR = 2.55, 95% CI = 1.85-3.53; p = 1.39 × 10-8; 1-year AUC = 0.768) and was externally validated (AUC = 0.719). Strong gene-level associations were observed between m6A regulators and metabolism-related genes, including WTAP-LDHA (r = 0.56) and FTO-ACACA (r = 0.75). GSEA identified significant enrichment of EMT (Epithelial-Mesenchymal Transition), TNF-α/NF-κB, IL-6/JAK/STAT3, angiogenesis, and hypoxia pathways, whereas glycolysis showed a non-significant enrichment trend (FDR = 0.294). Overall, these findings link m6A-associated transcriptional networks with metabolic and inflammatory features and prognostic variation in GBM, providing a hypothesis-generating framework for further experimental investigation.