Abstract
Objective: Gastric cancer (GC) is recognized as one of the prevailing solid malignant tumors globally, with a notable rate of recurrence and metastasis. Therefore, this study utilized database mining to analyze potential key genes (hub genes) that are associated with the progression and prognosis of GC, aiming to offer new clues for the prognosis and targeted treatment for GC.
Methods: This study utilized the GSE79973 dataset from the GEO to conduct DEGs in conjunction with the WGCNA. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was performed on disease-characteristic differentially expressed genes. In addition, a PPI network through the STRING database to screen for characteristic genes involved in the molecular mechanisms of GC. The diagnostic capabilities of these characteristic genes were ascertained through ROC curve analysis, integrating the clinical data of GC from TCGA.
Results: Systematic bioinformatics analysis pinpointed four genes—COL1A1, COL1A2, COL4A1, and TLR2—as closely related to the onset and progression of GC. ROC curve revealed their robust diagnostic and prognostic capabilities for GC (AUC(COL1A1)=0.9478, AUC(COL1A2)=0.8768, AUC(COL4A1)=0.8482, AUC(TLR2)=0.8452, all P < 0.0001), presenting significant clinical translational application value.
Conclusion: As newly discovered functional genes closely related to the onset and progression of GC, COL1A1, COL1A2, COL4A1, and TLR2, can be deemed as novel biomarkers for clinical diagnosis of GC, paving the way for new effective targets in the treatment of GC.