Identification of hub genes and its relationship with colorectal cancer patient prognosis

Abstract

Background Colon cancer is a common gastrointestinal malignancy that with high morbidity and mortality, however, predicting prognosis in colon cancer is imperfect, so searching for effective prognostic biomarkers is in urgent need. Our study aimed to investigate hub genes and their prognostic value in colon cancer via bioinformatics analysis. Methods The data set of colon cancer was downloaded from GEO database (GSE21510, GSE110225, GSE74602) for differential gene analysis using the GEO2R tool, and hub gene was identified by Cytoscape. The Human Protein Atlas database was used to analyze the expression of hub genes and evaluate their prognostic value using GEPIA. TIMER database was used to study the correlation between hub gene and immune infiltration of COAD cells. After that, PPI networks were established using the SRTING database and Cytoscape to analyze hub genes and protein-protein interactions. Finally, identify hub gene mutation through cBioPortal, the co-expression analysis of hub genes were identified by LinkedOmics, and gene set enrichment analysis (GSEA) were analyzed using the comprehensive functional classification database in the Web-based Web Gestalt. Results We identified 349 differentially expressed genes (DEGs), including 117 upregulated and 229 downregulated. The PPI network was constructed and four hub genes (AURKA, CCNB1, EXO1 and CCNA2) were selected by survival analysis and expression validation. The expression level of AURKA, CCNB1, EXO1 and CCNA2 were all higher in COAD tumor tissues than in adjacent tissues, as well as the immunohistochemical results. There has varying degrees gene mutation and varying degrees immune cell infiltration about AURKA, CCNB1, EXO1 and CCNA2, especially B cells and CD8 + T cells. GSEA was performed to determine co-expressed genes of AURKA, CCNB1, CCNA2 and EXO1, the results showed that they both had a close relationship with chromosome segregation, DNA replication, translational elongation and cell cycle. Conclusion Overexpression of AURKA, CCNB1, CCNA2 and EXO1 had a better prognosis in colon cancer and this consequences was correlation with gene mutation and infiltration of immune cells.