Explore the prognostic characteristics of immunogenic cell death- related genes in colon carcinoma based on multi-constraint canonical correlation analysis

Abstract

Abstract The prognosis for colon carcinoma is poor, it is a highly aggressive tumor, and its early symptoms are subtle. Hence, the need for more colon cancer prognostic indicators is urgent. Many long non-coding RNAs (lncRNAs) are known to actively contribute to the detection, prognosis, and treatment of tumors. On the other hand, immunogenic cell death (ICD) is vital for the prevention of tumors. This study seeks to research the mechanism of action of ICD-related lncRNAs in colon cancer and integrate ICD-related genes with lncRNAs. In order to investigate ICD-related lncRNAs, this research suggests a multi-constraint canonical correlation analysis. The prognostic model of the most significant lncRNAs is also constructed and verified in this paper using univariate and Lasso-Cox regression. In this study, colon cancer samples are split into high-risk and low-risk groups based on the model's risk score, and the impact of the model is assessed using the TCGA queue and the GEO queue, respectively. Also, we thoroughly examined the parallels and discrepancies in immune infiltration, functional enrichment, and medication sensitivity across the two risk groups. In conclusion, the ICD-related lncRNAs risk model developed in this research can serve as a guide for predicting the prognosis and therapeutic sensitivity of colon cancer patients.