Development and validation of a glycosyltransferase-associated prognostic model for melanoma and characterization of the tumor immune microenvironment using single-cell sequencing data

Abstract

This study aimed to develop a predictive model based on glycosyltransferase-related genes (GTs) to forecast the survival time of patients with Skin Cutaneous Melanoma (SKCM) and to explore the pathways and mechanisms through which GTs influence SKCM prognosis. Transcriptomic data of SKCM from The Cancer Genome Atlas (TCGA) were utilized for individualized predictive modeling, and the model's reliability was validated using GEO data. Univariate Cox regression and LASSO-Cox regression analyses were employed to select prognostically relevant biomarkers, and a predictive risk score was constr, ucted using multivariate Cox regression. Functional annotation of the risk score was performed through GO, KEGG, and GSEA analyses. The performance of the nomogram model was evaluated using ROC curves, calibration curves, and the concordance index (C-index). Furthermore, subsequent analyses based on risk grouping were conducted to assess immune infiltration, somatic mutations, and immune responses, and these findings were validated by real-time quantitative PCR (qPCR), Western Blot, and immunohistochemistry (IHC). Our results revealed a significant correlation between the risk score derived from multivariate Cox regression and the overall survival of SKCM patients. Enrichment analysis of the risk score indicated its association with immune functions. The nomogram model, which integrates the risk score with clinical prognostic factors, exhibited robust predictive performance in both training and validation datasets. Further analyses—including immune infiltration, single-cell analysis, somatic mutation analysis, and immune response assessment—demonstrated a strong correlation between the key gene MGAT4A and the infiltration of CD8⁺ T cells as well as monocytes/macrophages in tumor tissues. In summary, we have developed an individualized predictive model for forecasting the 1-year, 3-year, 5-year, and 10-year survival rates of SKCM patients. This model holds promise as a potential tool for guiding personalized diagnosis and treatment of SKCM.