Integrated single-cell and bulk RNA-seq for constructing a prognostic risk model of immune-related genes in colorectal cancer

Abstract

Abstract Colorectal cancer (COAD) ranks as the second leading cause of global cancer-related deaths. However, effective therapeutic strategies for advanced colorectal cancer remain limited. Despite the success of immunotherapy in cancer treatment, its applicability is significantly constrained by the heterogeneity of the tumor microenvironment. Therefore, the aim of this study is to develop an immune-related feature model at the single-cell level, categorize colorectal cancer patients based on gene transcription levels, and predict patients' prognosis, immune status, and treatment response. By analyzing single-cell transcriptomic data (scRNA-seq) of COAD, we identified eight cell types and selected immune cell marker genes (IRGs) for subsequent analysis. Utilizing these IRGs, we established an immune-related model to classify patients into high and low-risk groups for predicting overall survival (OS). The low-risk group exhibited high tumor mutation burden (TMB), increased immune activation, high microsatellite instability (MSI-H), longer overall survival (OS), and better response to immunotherapy. Conversely, the high-risk group displayed microsatellite stability (MSS), low TMB, immune suppression, and other characteristics. Additionally, we validated the model's performance in predicting immune treatment responses using external data from the IMvigor210 cohort. In summary, combining single-cell and bulk tissue transcriptome sequencing, we constructed a survival risk prognosis model that categorizes patients into high and low-risk groups. This model enables the prediction of patients' immune cell status, immune-related functions, and immunotherapy effectiveness. These findings provide valuable insights into the immune status, prognosis assessment, and the development of effective immunotherapeutic approaches for colorectal cancer.