Abstract
Glioblastoma (GBM), an aggressive primary tumor, is common in humans, accounting for 12-15% of all intracranial tumors, and has median survival of fewer than 15 months. Since a growing body of evidence suggests that conventional drugs are ineffective against GBM, our goal is to find emerging therapies that play a role in its treatment. This research constructs a risk model to predict the prognosis of GBM patients. A set of genes associated with GBM was taken from a GBM gene data bank, and clinical information on patients with GBM was retrieved from the Cancer Genome Atlas (TCGA) data bank. One-way Cox and Kaplan-Meier analyses were performed to identify genes in relation to prognosis. Groups were classified into high and low expression level of PTEN expression. Prognosis-related genes were further identified, and multi-factor Cox regression analysis was used to build risk score equations for the prognostic model to construct a survival prognostic model. The area under the ROC curve suggested that the pattern had high accuracy. When combined with nomogram analysis, GJB2 was considered an independent predictor of GBM prognosis. This study provides a potential prognostic predictive biological marker for GBM patients and confirms that GJB2 is a key gene for GBM progression.