Gastric cancer has a poor prognosis and high mortality. Cuproptosis, a novel programmed cell death, is rarely studied in gastric cancer. Studying the mechanism of cuproptosis in gastric cancer is conducive to the development of new drugs, improving the prognosis of patients and reducing the burden of disease.
The TCGA database was used to obtain transcriptome data from gastric cancer tissues and adjacent tissues. GSE66229 was used for external verification. Overlapping genes were obtained by crossing the genes obtained by differential analysis with those related to copper death. Eight characteristic genes were obtained by three dimensionality reduction methods: lasso, SVM, and random forest. ROC and nomogram were used to estimate the diagnostic efficacy of characteristic genes. The CIBERSORT method was used to assess immune infiltration. ConsensusClusterPlus was used for subtype classification. Discovery Studio software conducts molecular docking between drugs and target proteins.
We have established the early diagnosis model of eight characteristic genes (ENTPD3, PDZD4, CNN1, GTPBP4, FPGS, UTP25, CENPW, and FAM111A) for gastric cancer. The results are validated by internal and external data, and the predictive power is good. The subtype classification and immune type analysis of gastric cancer samples were performed based on the consensus clustering method. We identified C2 as an immune subtype and C1 as a non-immune subtype. Small molecule drug targeting based on genes associated with cuproptosis predicts potential therapeutics for gastric cancer. Molecular docking revealed multiple forces between Dasatinib and CNN1.
The candidate drug Dasatinib may be effective in treating gastric cancer by affecting the expression of the cuproptosis signature gene.