Development and validation of a disulfidptosis-related prognostic model for colorectal cancer using multi-omics analysis.

Abstract

This study aims to integrate multi-omic and clinical data concerning disulfidptosis-related genes (DRGs) to facilitate molecular typing and prognosis in colorectal cancer (CRC). Public databases provided CRC transcriptome and clinical data, enabling differential expression, genomic analyses, pathway enrichment, survival analysis, and subtyping based on the expression levels of 15 DRGs identified in published studies. Differentially expressed genes (DEGs) between subtypes were identified to create a disulfidptosis prognostic model using LASSO and Cox regression analyses. This model was evaluated by comparing risk scores, survival curves, cellular infiltration, and drug sensitivity between high- and low-risk groups. Analyses revealed differential expression, mutations, and copy number variations (CNV) in DRGs in CRC. Survival analysis demonstrated significant prognostic differences among DRG expression subtypes. GSVA and ssGSEA highlighted DRGs' regulatory roles in CRC. DEGs identified between DRG expression subtypes led to the classification into subtypes A and B. A disulfidptosis prognostic model, including genes VSIG4, SCG2, INHBB, DDC, CXCL13, KLK10, CXCL10, and CCL11A, was developed to stratify patients into high- and low-risk groups. This model displayed strong predictive capability (AUC = 0.700) and calibration. The risk score was also strongly associated with immune cell infiltration, stromal cell score, and stem cell index in the CRC tumor microenvironment. Drug sensitivity analysis indicated that high-risk samples were more responsive to most medications. We established a robust disulfidptosis prognostic model for CRC through comprehensive multi-omics analysis. Our findings provide valuable insights into the role of DRGs in CRC progression and disease management, presenting an important resource for further research.