Abstract
BACKGROUND: The breakthrough discovery of novel biomarkers with prognostic and diagnostic value enables timely medical intervention for the survival of patients diagnosed with gastric cancer (GC). Typically, in studies focused on biomarker analysis, highly connected nodes (hubs) within the protein-protein interaction network (PPIN) are proposed as potential biomarkers. However, this study revealed an unexpected finding following the clustering of network nodes. Consequently, it is essential not to overlook weakly connected nodes (nonhubs) when determining suitable biomarkers from PPIN. METHODS AND RESULTS: In this study, several potential biomarkers for GC were proposed based on the findings from RNA-sequencing (RNA-Seq) datasets, along with differential gene expression (DGE) analysis, PPINs, and weighted gene co-expression network analysis (WGCNA). Considering the overall survival (OS) analysis and the evaluation of expression levels alongside statistical parameters of the PPIN cluster nodes, it is plausible to suggest that THY1, CDH17, TGIF1, and AEBP1, categorized as nonhub nodes, along with ITGA5, COL1A1, FN1, and MMP2, identified as hub nodes, possess characteristics that render them applicable as biomarkers for the GC. Additionally, insulin-like growth factor (IGF)-binding protein-2 (IGFBP2), classified as a nonhub node, demonstrates a significant negative correlation with both groups within the same cluster. This observation underscores the conflicting findings regarding IGFBP2 in various cancer studies and enhances the potential of this gene to serve as a biomarker. CONCLUSION: The findings of the current study not only identified the hubs and nonhubs that may serve as potential biomarkers for GC but also revealed a PPIN cluster that includes both hubs and nonhubs in conjunction with IGFBP2, thereby enhancing the understanding of the complex behavior associated with IGFBP2.