Integrative machine learning and bioinformatics analysis unveil key genes for precise glioma classification and prognosis evaluation.

Gliomas exhibit significant heterogeneity and diverse molecular subtypes, and there are marked differences in treatment strategies and prognoses for gliomas of different grades and molecular types. However, current glioma molecular subtyping systems are still inadequate and often overlook the impact of the original tumour grade. This study focused on the differentially expressed genes between high-grade gliomas (HGGs) and low-grade gliomas (LGGs), aiming to construct a classification model that can be used to predict glioma prognosis. Through a comprehensive investigation involving differential expression analysis, weighted gene coexpression network analysis (WGCNA), proteinprotein interaction (PPI) network analysis, and univariate and multivariate survival analyses, we identified a core set of genes that influence glioma prognosis. Based on these core genes, we developed the novel malignancy prognosis gene score (MGP_Score) model and validated its stability and reliability with external datasets. This scoring system provides a new tool for assessing glioma prognosis. To explore the molecular feature differences between different prognostic subtypes in detail, we employed four machine learning algorithms: generalized linear model (GLM), random forest (RF), support vector machine recursive feature elimination (SVM-RFE), and eXtreme gradient boosting (XGB). With these algorithms, we successfully identified B2M, SRPX2, and SERPINH1 as specific diagnostic and prognostic biomarkers for the malignant subgroup. These biomarkers can not only effectively distinguish between HGG and LGG but also accurately predict patient survival rates. We not only revealed molecular differences among gliomas but also constructed a prognostic model, validated the effectiveness of the MGP_Score, and identified biomarkers with potential for clinical application. These findings provide a solid foundation and guidance for precision medicine for gliomas, with the potential to improve early diagnosis, personalized treatment, and prognosis assessment.