Abstract
BACKGROUND: Metabolic reprogramming and immune evasion synergistically drive breast carcinogenesis, but their combined impact remains unclear. METHODS: Transcriptomic data from the TCGA and GEO cohorts were integrated. Differentially expressed genes were identified, followed by WGCNA to detect immune-correlated co-expression modules. Immune-metabolism-related genes (IMGs) were screened using Genecards. Four machine learning algorithms (LASSO, SVM, RF, XGBoost) identified hub genes. The diagnostic value was evaluated by Kaplan-Meier and ROC analysis. CIBERSORT quantified immune microenvironment associations. The expression profiles of genes in different cells were plotted using single-cell RNA data. IHC validated protein expression in clinical samples. RESULTS: Research has found that SELENOP and PKMYT1 are key immune metabolic hubs. Compared with normal tissues, the expression of SELENOP was significantly decreased (p < 0.05), while PKMYT1 showed an upward trend (p < 0.05). Both of these genes have demonstrated high accuracy in the diagnosis of breast cancer and can effectively predict the overall survival period of patients. Low SELENOP expression is associated with high PKMYT1 expression levels, which is significantly related to changes in immune infiltration and the expression patterns of checkpoint proteins. Immunohistochemical detection further confirmed that these genes were significantly correlated with histological grade, LAG-3, CD244, ER, PR and Her-2 and other indicators (p < 0.05). CONCLUSION: SELENOP and PKMYT1 are novel immunomodulatory factors related to multiple pathological indicators of breast cancer and can be used as diagnostic biomarkers.