A Propionate Metabolism-Based Gene Signature Reveals Immunogenomic and Transcriptomic Determinants of Prognosis in Glioblastoma Through Multiomics Integration.

Glioblastoma (GBM) remains a lethal brain tumor with limited prognostic tools. Metabolic reprogramming, particularly in understudied pathways like propionate metabolism, may offer new biomarkers. Here, we identified a novel prognostic signature based on seven propionate metabolism-related genes (SLC9A1, ELANE, ACADS, SOAT2, MYD88, ADSL, and BMP2) from the TCGA-GBM cohort. A risk scoring model was constructed via LASSO Cox regression effectively stratified patients into high- and low-risk groups with significant survival differences, which was also validated in independent GEO datasets. Multiomics analysis revealed that the high-risk group was associated with an immunosuppressive microenvironment, characterized by increased immune checkpoint expression and distinct immune cell infiltration. Mutational profiling showed a strong association with key driver alterations, including enrichment of RB1 mutations in high-risk and IDH1 mutations in low-risk patients. Single-cell RNA-seq (scRNA-seq) analysis confirmed the specific enrichment of signature genes within malignant cells, and coexpression network analysis (hdWGCNA) further linked the high-risk phenotype to transcriptional modules. In conclusion, we established and validated a robust metabolic gene signature that not only predicts prognosis but also delineates a high-risk GBM subtype defined by integrated metabolic, immunogenomic, and transcriptional features, providing new insights into the determinants of GBM aggressiveness.