A Metabolism-Related Gene Signature Predicts the Prognosis of Breast Cancer Patients: Combined Analysis of High-Throughput Sequencing and Gene Chip Data Sets

Background and Aim: Hundreds of consistently altered metabolic genes have been identified in breast cancer (BC), but their prognostic value remains to be explored. Therefore, we aimed to build a prediction model based on metabolism-related genes (MRGs) to guide BC prognosis. Methods: Current work focuses on constructing a novel MRGs signature to predict the prognosis of BC patients using MRGs derived from the Virtual Metabolic Human (VMH) database, and expression profiles and clinicopathological data from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Results: The 3-MRGs-signature constructed by SER-PINA1, QPRT and PXDNL was found to be an independent prognostic factor for the survival of patients, and based on the model, the overall survival (OS) of the high-risk group was significantly lower. Furthermore, a nomogram was developed based on risk score and independent prognostic clinical indicators, and its validity of survival prediction was confirmed by the calibration curve, the concordance index, decision curve analysis and receiver operating characteristic curve. The ssGSEA analysis showed a negative correlation between immune cell infiltration and risk score, which is consistent with the GSEA result showing that low-risk score group was associated with activated immune processes. Half-maximal inhibitory concentration of chemotherapeutic drugs was estimated by pRRophetic algorithm to guide clinical medication. Conclusion: We constructed and validated an effective 3-MRGs (SERPINA1, QPRT and PXDNL)-based prognostic model, and demonstrated that lower-risk patients were associated with higher immune infiltrations, underscoring the importance of immune ecosystems in determining the prognosis of BC patients.