Abstract
Background: Gastric cancer (GC) is the most common malignant tumor. Due to the lack of practical molecular markers, the prognosis of patients with advanced gastric cancer is still poor. A number of studies have confirmed that the coagulation system is closely related to tumor progression. Therefore, the purpose of this study was to construct a coagulation-related gene signature and prognostic model for GC by bioinformatics methods. Methods: We downloaded the gene expression and clinical data of GC patients from the TCGA and GEO databases. In total, 216 coagulation-related genes (CRGs) were obtained from AmiGO 2. Weighted gene co-expression network analysis (WGCNA) was used to identify coagulation-related genes associated with the clinical features of GC. Last absolute shrinkage and selection operator (LASSO) Cox regression was utilized to shrink the relevant predictors of the coagulation system, and a Coag-Score prognostic model was constructed based on the coefficients. According to this risk model, GC patients were divided into high-risk and low-risk groups, and overall survival (OS) curves and receiver operating characteristic (ROC) curves were drawn in the training and validation sets, respectively. We also constructed nomograms for predicting 1-, 2-, and 3-year survival in GC patients. Single-sample gene set enrichment analysis (ssGSEA) was exploited to explore immune cells' underlying mechanisms and correlations. The expression levels of coagulation-related genes were verified by real-time quantitative polymerase chain reaction (qRT-PCR) and immunohistochemistry (IHC). Results: We identified seven CRGs employed to construct a Coag-Score risk model using WGCNA combined with LASSO regression. In both training and validation sets, GC patients in the high-risk group had worse OS than those in the low-risk group, and Coag-Score was identified as an independent predictor of OS, and the nomogram provided a quantitative method to predict the 1-, 2-, and 3-year survival rates of GC patients. Functional analysis showed that Coag-Score was mainly related to the MAPK signaling pathway, complement and coagulation cascades, angiogenesis, epithelial-mesenchymal transition (EMT), and KRAS signaling pathway. In addition, the high-risk group had a significantly higher infiltration enrichment score and was positively associated with immune checkpoint gene expression. Conclusion: Coagulation-related gene models provide new insights and targets for the diagnosis, prognosis prediction, and treatment management of GC patients.