Identification of Potential Diagnostic and Prognostic Biomarkers for Gastric Cancer Based on Bioinformatic Analysis.

Abstract

Gastric cancer (GC) ranks third for cancer-related fatalities worldwide. It is still unclear what causes GC to progress. Using integrated bioinformatics analysis, COL5A2 has been proved to be related to GC development, which may identify the likely pathogenic mechanism. Data from GC patients were gathered using The Cancer Gene Atlas (TCGA) and the gene expression omnibus (GEO). The level of COL5A2 expression was compared between paired GC and normal tissues. The differentially expressed genes (DEGs) in GC patients with high and low COL5A2 expression were identified using functional enrichment analysis to identify the signature pathways linked to the DEGs. The clinical pathologic traits connected to overall survival (OS) of GC patients were examined utilizing Cox regression and the Kaplan-Meier method. To assess the prognostic significance of COL5A2, receiver operating characteristic (ROC) curves was drawn. How the immune system infiltrate both normal gastric and GC tumor tissues was investigated. Using the human protein atlas (HPA) database, regression, and the Kaplan-Meier method, immunohistochemical analysis of DEG COL5A2 expression in GC tissues was carried out. The correlation between COL5A2 expression and the GC grouping was found to be highly significant. Functional annotations revealed that COL5A2 participates in extracellular matrix structure, collagen metabolism, and other biological processes (BPs). High COL5A2 expression was associated with poor prognostic and clinical features, such as clinical T, N, and M stages. ROC curves exhibited that COL5A2 might predict the occurrence of gastric cancer. The infiltration degree of 21 immune cell subsets, including activated dendritic cells (aDCs), CD8+ T cells, and cytotoxic cells, was found to be dramatically relevant to COL5A2. Immunohistochemical analysis indicated that the expression of COL5A2 in tumor tissues is higher than that in normal tissues. The COL5A2 gene may offer fresh perspectives on the pathogenic mechanism underlying GC, as well as potential biomarkers for estimating GC patient prognosis. As a result, COL5A2 may be a useful biomarker for predicting patient survival.