Construction of Glioma Prognosis Model and Exploration of Related Regulatory Mechanism of Model Gene

Objective: To analyze the prognostic role of differentially expressed genes (DEGs) in glioma. Methods: Kaplan-Meier (KM) and univariate Cox analysis were used to screen the common prognostic genes of LGG and GBM. Multivariate Cox regression analysis was included for further analysis. Prognostic risk models were constructed and the risk score of each patient was calculated. The HPA database was used to analyze the expression of model genes. GSCAlite platform was used to analyze model genes' regulatory network and drug sensitivity. TIMER analyzed the correlation between model genes and immune infiltration. Results: Multivariate Cox regression analysis showed that patient age, tumor grade, and patient risk score were independent risk factors for LGG prognosis. PTPRN and RGS14 were under-expressed in gliomas, and there was a synergistic effect on TSC/mTOR but inhibited RAS/MAPK, hormone AR and ER pathways in LGG. Over-expressed MTHFD2 and HOXB2 showed antagonistic effects with PTPRN and RGS14. Afatinib, gefitinib, trametinib, methotrexate, FK866 and vorinostat were more sensitive to model genes. The expression of FERMT1, HOXB2, and PTPRN was significantly correlated with the immune infiltration level of LGG. Conclusions: The prognostic risk model, molecular mechanism, and regulation of model genes play an important role in glioma.