Comprehensive analysis reveals independent prognostic biomarkers in glioblastoma

Background: Glioblastoma (GBM) is a highly malignant intracranial tumor with an extremely poor prognosis. Screening and identification of potential prognostic biomarkers are essential for clinical treatment of GBM. Methods In this study, a total of 199 differentially expressed genes (DEGs) were identified based on GSE4290, GSE19728 and GSE50161 datasets. Function enrichment analyses and protein-protein interaction (PPI) network were performed. As a result, 10 genes (CDK1, TOP2A, RRM2, DTL, PBK, NUSAP1, NDC80, DLGAP5, KIF14, FOXM1) were selected as hub genes using CytoHubba in Cytoscape. Subsequently, GEPIA, Human Protein Atlas, PROGgeneV2 and cBioPortal databases were used to study the roles of hub genes in the prognosis of GBM. Results The results showed that 10 hub genes were over-expression in GBM patients and higher expressions of hub genes were significantly correlated with shorter overall survival (OS) in GBM patients. Multivariate analysis indicated that RRM2, DLGAP5 and KIF14 were independent prognostic factors of GBM patients. Three independent prognostic biomarkers were also found to be correlated with immune cell infiltration and clinical features of glioma patients. Conclusion Our results indicated that RRM2, DLGAP5 and KIF14 could be independent prognostic biomarkers of GBM patients.