The use of mRNA vaccines to treat colon cancer (CC) remains largely unexplored. The aim of this study is to identify CC antigens that may potentially be used for the development of anti‐CC mRNA vaccines and to select suitable patients for vaccination. Immune subtypes are identified by consensus clustering analysis, and the visualization of the CC immune landscape is performed by graph‐learning‐based dimensionality reduction. Plot cell trajectory and the WGCNA are applied to resolve the heterogeneity of the population suitable for vaccination. Four tumor antigens, COQ2, BGN, CASP6, and FAP, which are related to prognosis and infiltration of antigen‐presenting cells, are identified in CC. CC patients are divided into four immune subtypes characterized by differential molecular, cellular, pathological, and clinical features. Group B has the characteristics of immunologically “cold” tumors, and the immunosuppressive phenotype has worse survival than other groups. Moreover, different immune subtype tumors differed in the expression of immune checkpoints and immunogenic cell death modulators. Last, immune cell components in individual patients are revealed by the immune landscape. COQ2, BGN, CASP6, and FAP are potential antigens for use in a CC‐mRNA vaccine and suitable for patients in Group B.