Identification of molecular subtypes for breast cancer based on butyrate metabolism-related genes to assess prognosis and immune landscape.

Background: Breast cancer (BRCA) ranks among the highest commonly occurring malignant tumors globally, posing a significant risk to women's health. Numerous studies suggest that butyrate holds potential as an anti-cancer compound across various human cancers. However, the impact on the initiation and progression of BRCA remains insufficiently explored. Therefore, this study aimed to identify molecular subtypes of BRCA based on butyrate metabolism-related genes, construct a prognostic model, and explore the associated immune landscape to provide insights for prognosis assessment and therapeutic strategies. Methods: Transcriptomic data and clinical details of BRCA patients were obtained from The Cancer Genome Atlas (TCGA) database. The patients were categorized into two distinct subtypes utilizing the K-means method. A predictive model was constructed employing the least absolute shrinkage and selection operator (LASSO), random forest, and multivariate Cox regression methods. To assess the model's performance, Kaplan-Meier survival analysis and receiver operating characteristic (ROC) curves were employed. Additionally, calibration and decision curve analysis (DCA) were utilized to further evaluate its accuracy. A predictive model in the form of a nomogram was designed to estimate the prognosis of BRCA. Immune scores in tumor tissues were assessed using the ESTIMATE algorithm. Single-sample gene set enrichment analysis (ssGSEA) was applied to examine the immune microenvironment. Tumor mutational burden (TMB) analysis was carried out to evaluate the mutation frequency in genes. Additionally, a drug sensitivity assessment was carried out. As a final step, we employed quantitative real-time polymerase chain reaction (qRT-PCR) assays to experimentally validate the expression of the characterized genes within BRCA samples. Results: We categorized the patients into two separate subtypes, then eight signature genes were functioned as key biomarkers for prognosis. Individuals in the high-risk group experience reduced survival rates. The high-risk groups exhibited a lower immune cell infiltration patterns and immune checkpoint molecules. To conclude, we identified candidate drugs and assessed their sensitivity for BRCA treatment. Conclusions: In conclusion, we established a prognostic model for BRCA with eight signature genes, and these results could offer new targets for BRCA therapy.