A novel cancer-associated fibroblast-related gene signature for predicting diffuse large B cell lymphoma prognosis using weighted gene co-expression network analysis and machine learning.

Abstract

ObjectiveOur objective was to investigate a novel cancer-associated fibroblast-related gene signature for predicting clinical outcomes in patients with diffuse large B cell lymphoma.MethodsThe cancer-associated fibroblast-related module genes were identified from Gene Expression Omnibus datasets using weighted gene co-expression network analysis in our retrospective study. Least Absolute Shrinkage and Selection Operator Cox regression was applied to screen a minimal set of genes and construct a prognostic cancer-associated fibroblast-related gene signature for diffuse large B cell lymphoma. Kaplan-Meier plots and receiver operating characteristic curves were used to assess the prognostic performance of the prognostic cancer-associated fibroblast-related genes. A nomogram encompassing the clinical information and prognostic scores of the patients was constructed. Additionally, the relationships of the gene signature with the immune landscape and drug sensitivity were explored.ResultsCapitalizing on machine learning, we developed a prognostic cancer-associated fibroblast-related gene signature risk model, efficiently categorizing patients with diffuse large B cell lymphoma into high- and low-risk groups and exhibiting a more robust capacity for survival prediction. The nomogram showed stronger prognostic ability than the clinical factor-based model or the risk score alone. We also observed significant differences in immune cell profiles and therapeutic responses between the two groups, offering valuable insights for developing personalized treatments for diffuse large B cell lymphoma.ConclusionsWe developed a prognostic cancer-associated fibroblast-related gene-based genetic risk model to predict the prognosis of diffuse large B cell lymphoma, potentially aiding in treatment selection.