Integrated Network Pharmacology, Machine Learning and Experimental Validation to Identify the Key Targets and Compounds of

Background: TiaoShenGongJian (TSGJ) decoction, a traditional Chinese medicine for breast cancer, has unknown active compounds, targets, and mechanisms. This study identifies TSGJ's key targets and compounds for breast cancer treatment through network pharmacology, machine learning, and experimental validation. Methods: Bioactive components and targets of TSGJ were identified from the TCMSP database, and breast cancer-related targets from GeneCards, PharmGkb, and RNA-seq datasets. Intersection of these targets revealed therapeutic targets of TSGJ. PPI analysis was performed via STRING, and machine learning methods (SVM, RF, GLM, XGBoost) identified key targets, validated by GSE70905, GSE70947, GSE22820, and TCGA-BRCA datasets. Pathway analyses and molecular docking were performed. TSGJ and core compounds' effectiveness was confirmed by MTT and RT-qPCR assays. Results: 160 common targets of TSGJ were identified, with 30 hub targets from PPI analysis. Five predictive targets (HIF1A, CASP8, FOS, EGFR, PPARG) were screened via SVM. Their diagnostic, biomarker, immune, and clinical values were validated. Quercetin, luteolin, and baicalein were identified as core components. Molecular docking confirmed their strong affinities with predicted targets. These compounds modulated key targets and induced cytotoxicity in breast cancer cell lines in a similar way as TSGJ. Conclusion: This study reveals the main active components and targets of TSGJ against breast cancer, supporting its potential for breast cancer prevention and treatment.