RESUMO
Objective To identify immune-related molecular markers in an attempt to predict prognosis of colon adenocarcinoma (COAD). Methods Immune related genes (IREGs) was analyzed based on the TCGA database. Weighted gene co-expression network analysis (WGCNA) and Cox regression analysis were used to establish risk models. According to the median risk score, COAD patients were divided into high risk and low risk groups. The prognostic difference were compared between the two groups. The function of the model was validated using GEO. Results A total of 1015 IREGs was obtained. The established model consisted of three genes: RAR related orphan receptor C (RORC), leucine-rich repeat Fli-I-interacting protein 2 (LRRFIP2) and lectin galactoside-binding soluble galectin 4 (LGALS4). The high-risk group had significantly poorer prognosis than low-risk group in the GEO database, and it was validated using a GEO database. Further analysis via univariate and multivariate Cox regression analyses revealed that risk model could function as independent prognostic factor for COAD patients. Conclusion The risk model based on IREGs can predict the prognosis of patients with COAD.