Fig. 3
Construction and functional annotation of the m6A gene characteristic model, construction of prognostic m6A feature model, and regulation of BPs in patients with CESC. A Based on the expression characteristics of DEGs between high and low m6A risk scores, unsupervised analysis and hierarchical clustering were performed to classify patients into three categories: Geneclusters A, B, and C, and DEGs were divided into signature -A and -B gene sets according to their expression changes. B Survival analysis showed significant differences in prognosis among the different Genecluster groups, and Genecluster A had the worst prognosis. C GO analysis showed that signature gene-A was involved in leukotriene catabolic processes, leukotriene B4 catabolic processes, and leukotriene B4 metabolic processes. D KEGG analysis showed that the signature-A gene set is closely related to glycosphingolipid biosynthesis-lacto and neolacto series, nicotine addiction, and other pathways. E GO analysis showed that the signature-B gene set was involved in heat generation, positive regulation of angiogenesis, and positive regulation of vasculature development. F KEGG enrichment analysis showed that signature-B and age-range signaling pathways in diabetic complications, pertussis, cellular senescence, and other pathways are closely related. G Sankey plots show correlations between the different gene clusters (Geneclusters A, B, and C, prognostic m6A traits (m6A group), and patient prognostic status (OS). H Survival analysis showed that the prognostic m6A feature model could better predict the OS of patients with CESC (log-rank P < 0.001). I, J GSEA for high- and low-risk patients. A signature gene set was downloaded from the MSigDB database, with 1,000 replicates per run. Epithelial-mesenchymal transition, angiogenesis, and Myc targets V1 were mainly enriched in high-risk groups, while bile acid metabolism, KRAS signaling DN, estrogen response late, and other pathways were significantly enriched in low-risk patients