The Construction of a Novel Ferroptosis-Related LncRNA Model to Predict Prognosis in Colorectal Cancer Patients

Colorectal cancer (CRC) is the most common gastrointestinal tumor with poor prognosis. Ferroptosis is a pivotal form of programmed iron-dependent cell death different from autophagy and apoptosis, and long non-coding RNA (lncRNA) can influence the prognosis of CRC via regulating ferroptosis. Nowadays, artificial intelligence (AI) driven solutions, especially machine learning (ML) methodologies, are becoming effective tools in increasing the likelihood of the development of new prognostic and predictive biomarkers of diseases. In this study, with ML-associated tools, a prognostic model was constructed and validated by screening ferroptosis-related lncRNAs associated with prognosis based on the transcriptome data and survival data of CRC patients in TCGA database. Regarding the established prognostic models, differences in signaling pathways and immune infiltration, as well as differences in immune function, immune checkpoints, and m6A-related genes were also analyzed. We obtained a total of 6 ferroptosis-related lncRNAs, and found that the prognostic model could accurately predict the prognosis of CRC patients. Significant differences were found in multiple signaling pathways, as well as immune infiltration, immune function, immune checkpoints, and m6A-related genes between high and low risk groups. Our study provides an efficient prediction tool for CRC patients and contributes to guide the personalized treatment.