RNA modification patterns based on major RNA modifications define tumor microenvironment characteristics in glioblastoma.

RNA modifications play a major role in tumorigenicity and progression, but the expression and function in glioblastoma (GBM) have not been well described. In this study, we developed a GBM score based on the differentially expressed genes (DEGs) between groups showing RNA modification patterns. We assessed the association between the GBM score and tumor microenvironment (TME) characteristics. Based on the gene expression of these regulators, we identified two clusters with distinct RNA modification patterns. Kaplan-Meier survival curves showed that patients in cluster 1 had worse survival than those in cluster 2. Kaplan-Meier and multivariate Cox regression analyses showed that GBM scores (based on DEGs between RNA modification patterns) are an independent predictive biomarker for patient prognosis. Besides, we found that samples with high scores were significantly associated with epithelial-to-mesenchymal transition and immune checkpoints, while samples with low scores were associated with cell cycle regulation. Importantly, GBM-score markedly positively correlated drug resistance, while negatively correlated with drug sensitive. The responders of anti-PD-1/PD-L1 immunotherapy tend to have a lower GBM score than non-responders. In conclusion, our comprehensive analysis of multiple RNA modifications in GBM revealed that RNA modification regulators were closely correlated with TME.