Abstract
Glioblastoma (GBM) is the most aggressive primary brain cancer, with poor prognosis due to its aggressive behavior and high heterogeneity. This study aimed to identify cellular senescence (CS) and lipid metabolism (LM)-related prognostic genes to improve GBM prognosis and treatment. Transcriptome and scRNA-seq data, CS-associated genes (CSAGs), and LM-related genes (LMRGs) were acquired from public databases. Prognostic genes were identified by intersecting CSAGs, LMRGs, and differentially expressed genes (DEGs), followed by WGCNA and univariate Cox regression. A risk model and nomogram were constructed. Analyses covered clinicopathological features, immune microenvironment, somatic mutations, and drug sensitivity. GBM scRNA-seq data identified key cells and prognostic gene expression. SOCS1 and PHB2 were identified as prognostic markers, contributing to the construction of a robust risk model with excellent predictive ability. High-risk group (HRG) patients had poorer survival, higher immune and stromal scores, and distinct somatic mutation profiles. Drug sensitivity analysis revealed significant differences in IC50 values. In microglia differentiation, SOCS1 and PHB2 showed dynamic expression patterns. These findings provide new strategies for GBM prognosis and treatment.