The Circadian Rhythm Gene Network Could Distinguish Molecular Profile and Prognosis for Glioblastoma

Increasing evidence highlights the role of aberrant circadian rhythm gene expression in glioblastoma (GBM) progression, but the impact of the circadian rhythm gene network on GBM molecular profiles and prognosis remains unclear. A total of 1042 GBM samples from six public datasets, TCGA and CGGA, were analyzed, with GBM samples stratified into three circadian core-gene patterns using unsupervised clustering based on the expression profiles of 17 circadian rhythm genes. The Limma R package identified differentially expressed genes (DEGs) among the three patterns, and a secondary clustering system, termed circadian-related gene pattern, was established based on DEGs. A circadian risk score was constructed using the Least Absolute Shrinkage and Selection Operator (LASSO) regression algorithm, and the efficiency of these patterns and the circadian risk score in distinguishing molecular profiles and predicting prognosis was systematically analyzed. The relationship between the circadian risk score and response to immune or targeted therapy was examined using the GSE78200 and IMvigor210 datasets. The results showed that GBM patients were clustered into three circadian core-gene patterns based on the expression profiles of 17 core circadian genes, with distinct molecular profiles, malignant characteristics, and patient prognoses among the patterns. Thirty-two DEGs among these patterns were identified and termed circadian-related genes, and secondary clustering based on these 32 DEGs classified GBM samples into two circadian-related gene patterns, which also predicted molecular profiles and prognosis. A circadian risk scoring system was established, allowing the calculation of individual risk scores based on the expression of 10 genes, where GBM patients with lower circadian risk scores had prolonged overall survival and less aggressive molecular subtypes, while higher circadian risk scores correlated with better responses to MAPK-targeted therapy. In conclusion, this study established two clustering patterns based on 17 circadian rhythm genes or 32 circadian-related genes, enabling the rapid classification of GBM patients with distinct molecular profiles and prognoses, while the circadian risk scoring system effectively predicted survival, molecular profiles, and therapeutic responses for individual GBM patients, demonstrating that the circadian rhythm gene network can distinguish molecular profiles and prognosis in GBM.