Single-cell multi-omics uncovers CPS1 as a breast cancer immune evasion therapeutic target.

Despite significant advances in early detection and therapeutic interventions, breast cancer persists as the most frequently diagnosed malignancy and the leading cause of cancer-related deaths among women globally. Although multiple prognostic signatures have been proposed, their predictive power and clinical applicability remain limited. In this study, we utilized an integrated approach combining single-cell multi-omics analysis with machine learning to comprehensively examine the clinical relevance of mitochondrial-related gene sets in TCGA-BRCA and developed a mitochondrial gene set scoring system, termed MitoScore. Based on the median of MitoScore, BRCA patients were classified into high-risk and low-risk groups. Our multi-omics analysis revealed that BRCA patients with higher Mitoscore exhibited poorer prognoses compared to those with lower MitoScore. The predictive ability of the model was successfully validated using an external GEO dataset. Immune infiltration analysis further indicated that high-risk group contributed to an immunosuppressive tumor microenvironment, marked by a decrease in CD8