Abstract
Cervical cancer (CESC) is a gynecologic malignant tumor associated with high incidence and mortality rates because of its distinctive management complexity. Herein, we characterized the molecular features of CESC based on the metabolic gene expression profile by establishing a novel classification system and a scoring system termed as METAscore. Integrative analysis was performed on human CESC samples from TCGA dataset. Unsupervised clustering of RNA sequencing data on 2,752 formerly described metabolic genes identified three METAclusters. These METAclusters for overall survival time, immune characteristics, metabolic features, transcriptome features, and immunotherapeutic effectiveness existed distinct differences. Then we analyzed 207 DEGs among the three METAclusters and as well identified three geneclusters. Correspondingly, these three geneclusters also differently expressed among the aforementioned features, supporting the reliability of the metabolism-relevant molecular classification. Finally METAscore was constructed which emerged as an independent prognostic biomarker, related to CESC transcriptome features, metabolic features, immune characteristics, and linked to the sensitivity of immunotherapy for individual patient. These findings depicted a new classification and a scoring system in CESC based on the metabolic pattern, thereby furthering the understanding of CESC genetic signatures and aiding in the prediction of the effectiveness to anticancer immunotherapies.