Metabolic reprogramming and prognostic modeling in pancreatic cancer: insights from WGCNA.

Abstract

Purpose: Metabolic reprogramming plays a crucial role in multiple malignant features of pancreatic cancer (PC). However, few studies have comprehensively examined metabolic features of PC and provided guidance for their treatment.

Methods: This study tried to identify metabolism-associated hub genes based on metabolic phenotypic levels through weighted gene co-expression network analysis, and constructed a risk model for PC, then verified its accuracy and explored the potential mechanisms.

Results: We screened out five metabolic hub and prognostic genes (DLX3, HMGA2, SPRR1B, MYEOV, and FAM111B) and constructed a novel metabolism-associated gene signature to predict the prognosis of PC. The model was verified efficacy and demonstrated with good performance through analysis of Kaplan-Meier plotter, receiver operating characteristic curves, comparing with reported models, application in predicting drug sensitivity and constructing a nomogram model. Correlation analysis revealed a close association between the levels of risk score and DNA damage response (DDR, correlation coefficient: 0.41, P < 0.001). Enrichment analysis indicated that risk scores were derived from multiple metabolic or proliferative pathways, providing further evidence that metabolism may mediate DDR to affect PC survival.

Conclusion: Through bioinformatics analysis, we identified five prognostic relevant differentially expressed genes highlighting the role of metabolism-associated factors in pancreatic cancer, which reveals a strong correlation ship with DDR, offering new insights into treatment strategies that combine metabolism with DDR.