Identification of Biomarkers Associated with the Prognoses of Colorectal Cancer Patients

Colorectal cancer (CRC) is a common cancer. As metastasis and recurrence are main causes of CRC death, it is of great significance to find prognostic biomarkers.Data related to CRC were collected from GEO database. The patients were grouped based on clinical information, and the differentially expressed genes (DEGs) were obtained by differential analysis. GO and KEGG pathway enrichment analyses were conducted based on DEGs. Cox combined with LASSO regression analysis was applied to screen out the key genes that used to build the prognostic model. Survival curve and receiver operating characteristic curve were employed to evaluate the validity and reliability of the model. Cox regression analysis was applied to determine the independence of risk score. GSEA and GSVA analyses were performed on patients with different risks according to the risk model scores, and the prognostic nomogram was plotted combined with clinical data. Also, qRT-PCR was applied to examine the expression status of the screened signatures in clinical cases.We obtained 302 DEGs by dividing CRC patients into early-stage and advanced-stage groups. The results of enrichment analyses demonstrated that the DEGs were mainly concentrated in tissues of extracellular matrix, epithelial cell proliferation, and cell adhesion-related pathways. Regression identified 9 hub genes notably correlated with prognosis, including CLK1, SLC2A3, LIPG, EPHB2, ATOH1, PLCB4, GZMB, CKMT2, and CXCL11. The validation of the risk model proved that the risk model was accurate and could independently determine the prognosis of patients. Finally, differences were found in pathway activity of extracellular matrix secretion, plaque secretion, Notch signaling pathway, and tight junctions in high-risk and low-risk patients. In addition to LIPG and CKMT2, other feature genes were notably overexpressed in CRC tumor tissues.The results proved that the expression levels of the 9 biomarkers could be used to predict the prognosis of CRC patients.