Prognostic significance of glycosylation-related genes as risk markers in colon cancer

ObjectiveA frequent post-translational alteration of proteins called glycosylation has been strongly linked to the development and progression of cancer. Targeting glycosylation may improve cancer treatment outcomes. This study intended to investigate relationship and prognostic significance of glycosylation-related gene set features with colon cancer survival, immunity, drug sensitivity, etc.Methods524 colon cancer patients were included from TCGA database as training cohort. GSE29621 was external validation cohort. Univariate analysis, LASSO, multivariate regression analysis, K-M survival curve, and ROC curve analysis were used to construct and validate a glycosylation-related gene-based riskscore prognostic model. The CIBERSORT method and TIDE algorithm were utilized to analyze and evaluate differences in immune levels between high- and low-risk groups and their response to immunotherapy. Based on the DSigDB database, potential drugs with potential targeting effects on the prognostic model genes were predicted.ResultsAccording to a series of regression analyses, we constructed a prognostic model with six glycosylation genes. The model showed favorable prognostic prediction ability in both training and validation sets. Relevant to high-risk group, low-risk group presented better survival rates, higher immune cell infiltration levels, lower TIDE scores, and a higher proportion of patients with potential response to immunotherapy. In addition, potential anti-tumor drugs such as 67526-95-8, verteportin, uracil, Pemetrexed disodium, and fisetin were screened through the DSigDB database.ConclusionIn summary, a validated prognostic model for colon cancer was constructed with glycosylation genes. The model could act as an independent prognostic factor for colon cancer.