Objective: Globally, one of the main causes of cancer-related mortality is Colon Adenocarcinoma (COAD). In this study, a new special Immune Cell Functions (ICF) risk model was constructed using single-cell and bulk RNA sequencing data to develop a new understanding and clinical applications for COAD. Methods: The immune function gene sets were downloaded from a literature reference, and the COAD single-cell dataset GSE146771 was downloaded from the Tumour Immune Single Cell Hub database. Using Lasso analysis, a multiple gene signature was made from the enrichment scores of immune function gene sets that were enriched in different ways. Robust validation of the signature was then performed in multiple independent cohorts. After that, we built the model using a 10-fold cross-test and evaluated its independence for clinical usage using a nomogram. We also investigated the connection between signature and immune function, genetic variation, immunotherapy, and the cancer immunological microenvironment. Lastly, we used qPCR and immunohistochemistry to examine the expression of the unreported model genes. To find the regulatory functions of unreported model genes, an EdU assay was employed. Results: First, 20 differentially enriched immune function gene sets were identified. Ten genes can be used as a risk profile to assess the prognosis of colon cancer, according to Lasso regression analysis. Signature performance was stable in both the training cohort and two independent GEO external cohorts, and risk scores were confirmed as independent prognostic factors. At the same time, our risk model continued to be highly predictive across various clinical clusters and clinical characteristics, such as immune checkpoints, tumour genome mutations, and chemotherapeutic drug resistance. Patients in the low-risk group have exhibited a higher chance of benefiting from immunotherapy, according to immunotherapy response research. qPCR and immunohistochemistry analysis have revealed SIRP5 expression as high in COAD tissues, while CMC1 and ASAH1 expression has been found to be low. According to the findings of the functional experiment, SIRP5, CMC1, and ASAH1 may control the ability of CRC cells to proliferate. Conclusion: In this study, using scRNA-seq and bulk RNA-seq data, we created a risk model to predict the prognosis and effectiveness of immunotherapy in patients with COAD. In addition, we have discovered three model genes (SIRP5, CMC1, and ASAH1) that have not been reported before. These genes have the potential to be novel therapeutic targets in Colorectal Cancer (CRC). These findings suggest that this model could be used to evaluate the prognostic risk and identify potential targets for COAD patient treatment.