Construction of taurine metabolism-related risk model and subtype identification in colorectal cancer to predict prognosis and immunological features.

Background: Colorectal cancer (CRC) represents a significant global health burden. Despite advances in treatment, survival rates remain low for patients diagnosed at advanced stages. Although taurine plays an indispensable role in cancer development and progression, its specific relationship with CRC prognosis has not been fully explored. Therefore, this study aims to construct a prognostic model based on taurine metabolism-related genes (TMRGs) to evaluate its potential value in CRC risk stratification, tumor microenvironment remodeling, and the development of personalized treatment strategies. Methods: This study integrated multi-omics data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases to systematically screen differentially expressed TMRGs (DETMRGs). Using least absolute shrinkage and selection operator (LASSO) regression and multivariate Cox analysis, we constructed a prognostic model and established a risk stratification framework. Molecular subtyping was performed based on these signature genes, followed by a comprehensive evaluation of the immune microenvironment (CIBERSORT/ESTIMATE), tumor mutational burden, drug sensitivity (pRRophetic/CellMiner), and immunotherapy response [immunophenoscore (IPS)/IMvigor210 cohort] to thoroughly validate the clinical predictive utility of the model. Results: This study identified 175 DETMRGs and constructed a prognostic model comprising 9 key genes. The model demonstrated robust survival predictive capability in both training and independent validation cohorts [area under the receiver operating characteristic curve (AUC) >0.74], effectively stratifying patients into molecular subtypes with significant survival disparities. In-depth analyses revealed that the high-risk group was enriched in pro-oncogenic signaling pathways (e.g., WNT, RTK-Ras) and exhibited an immunosuppressive microenvironment (elevated Treg infiltration, reduced immune scores) with inferior response to programmed cell death ligand 1 (PD-L1) immunotherapy. Conversely, the low-risk group activated immune-related pathways (e.g., JAK-STAT) and showed heightened sensitivity to chemotherapy agents such as 5-fluorouracil (5-FU). Conclusions: The taurine metabolism-related risk model developed in this study reliably predicts the prognosis of CRC patients and their response to immunotherapy, thereby enabling personalized treatment strategies.