Abstract
Glioma is the primary tumor with the highest incidence and the worst prognosis in the human central nervous system. Epithelial-mesenchymal transition (EMT) and immune responses are two crucial processes that contribute to it having the worst prognosis. However, a comprehensive correlation between these two processes remains elusive. The mRNA expression profiles and corresponding clinical data of patients with glioma were downloaded from public databases. EMT-related genes were collected and provided in the dbEMT database. Risk scores, Lasso regression, and enrichment analysis were conducted for functional validation. In our study, we used unsupervised clustering of EMT gene expression profiles to classify gliomas into two subtypes. We assessed the reliability of this classification system by testing it in three independent cohorts. Each subtype had different clinical and immune system characteristics. The study suggests a possible link between EMT and immune responses in gliomas.