PurposeColon adenocarcinoma (COAD) is the most common type of colorectal cancer (CRC) and is associated with poor prognosis. Emerging evidence has demonstrated that glycosylation by long noncoding RNAs (lncRNAs) was associated with COAD progression. To date, however, the prognostic values of glycosyltransferase (GT)-related lncRNAs in COAD are still largely unknown.MethodsWe obtained the expression matrix of mRNAs and lncRNAs in COAD from The Cancer Genome Atlas (TCGA) database. Then, the univariate Cox regression analysis was conducted to identify 33 prognostic GT-related lncRNAs. Subsequently, LASSO and multivariate Cox regression analysis were performed, and 7 of 33 GT-related lncRNAs were selected to conduct a risk model. Gene set enrichment analysis (GSEA) was used to analyze gene signaling pathway enrichment of the risk model. ImmuCellAI, an online tool for estimating the abundance of immune cells, and correlation analysis were used to explore the tumor-infiltrating immune cells in COAD. Finally, the expression levels of seven lncRNAs were detected in colorectal cancer cell lines by reverse transcription-quantitative polymerase chain reaction (RT-qPCR).ResultsA total of 1,140 GT-related lncRNAs were identified, and 7 COAD-specific GT-related lncRNAs (LINC02381, MIR210HG, AC009237.14, AC105219.1, ZEB1-AS1, AC002310.1, and AC020558.2) were selected to conduct a risk model. Patients were divided into high- and low-risk groups based on the median of risk score. The prognosis of the high-risk group was worse than that of the low-risk group, indicating the good reliability and specificity of our risk model. Additionally, a nomogram based on the risk score and clinical traits was built to help clinical decisions. GSEA showed that the risk model was significantly enriched in metabolism-related pathways. Immune infiltration analysis revealed that five types of immune cells were significantly different between groups, and two types of immune cells were negatively correlated with the risk score. Besides, we found that the expression levels of these seven lncRNAs in tumor cells were significantly higher than those in normal cells, which verified the feasibility of the risk model.ConclusionThe efficient risk model based on seven GT-related lncRNAs has prognostic potential for COAD, which may be novel biomarkers and therapeutic targets for COAD patients.