As the most prevalent brain tumor, glioma is malignant with poor prognostic outcomes. As a result, it is of great importance to detect biomarkers for glioma diagnosis and prognosis. In this study, we determined grade-based characteristic gene clusters with gradual expression following grade change, including 1479 down- and 526 up-regulated genes. Combined interaction among proteins originating from these genes was analyzed, and hub genes were exhibited after GSEA enrichment, containing 12 and 11 genes which were correlated with prognostic outcomes, named as unfavorable and favorable gene sets, respectively. The GSVA score of each gene set was calculated and divided into high/low groups; we observed that cases in the low score group had better outcomes than the high score group based on the GSVA of the unfavorable set, while with favorable GSVA score, the low group had poorer outcomes than the high group. Eventually, we compared a variety of infiltrating immune cells between low/high GSVA subgroup, showing various immune cell types (B cell naive, activated mast cells, resting CD4 memory T cell, and so on) with opposite proportion. And interestingly, these cell types also accounted for a contrary percentage between unfavorable and favorable conditions. In conclusion, these two hub gene sets are of good importance as an evaluation system for clinical grade classification and prognosis prediction.
Keywords: GSVA score; Glioblastoma; Glioma; Molecular stratification; Prognosis prediction.
© 2022. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.