Identification of a 5-gene-based signature to predict prognosis and correlate immunomodulators for rectal cancer

Abstract

Background: Specific tumor markers have yet to be identified in rectal cancer. This study aims to identify a novel genetic signature in rectal cancer to provide clues for survival and immunotherapy. Methods: DEGs were obtained from two GEO datasets of rectal cancer. By using data from TCGA and GSE133057, two cohorts of rectal cancer were applied to establish and evaluate the signature. A nomogram was constructed for training and validation. We integrated the risk-score with clinicopathological features and assessed its interplay with immune cells and molecules. Finally, our study performed functional annotations, gene-targeted miRNAs, and single-cell analysis. Results: A total of 468 DEGs were identified, and a signature consisting of 5 genes (CLIC5, ENTPD8, PACSIN3, HGD, and GNG7) was selected to calculate the risk-score. The model exhibited high performance in time-dependent ROC and a nomogram. Further results showed that overall survival was significantly worse in the high-risk group. As an independent prognostic factor, the risk-score was associated with vascular invasion. There was a dramatic difference in nonregulatory CD4+ and CD8+ T cells between the high and low-risk groups, and the 5 genes were correlated with immune inhibitors. There was a considerable difference in autophagy, immune, cell cycle, infection, and apoptosis-associated terms and pathways in GO and KEGG. The functional states of differentiation, apoptosis, and quiescence were closely related to the 5-gene signature in single-cell analysis. Conclusion: Our results suggest that the signature could serve as a novel prognostic biomarker in rectal cancer, which might benefit decision-making regarding immunotherapy.

Keywords

• Rectal cancer
• Prognosis
• Bioinformatics
• Differentially expressed genes
• Biomarker