Bioinformatics and network pharmacology-based investigation of matrine’s therapeutic roles in breast cancer

Breast cancer (BC) stands out as a leading malignancy in female populations, exhibiting a gradual global increase in cases. Matrine (MAT), a bioactive alkaloid derived from traditional Chinese medicine (TCM), has demonstrated anticancer potential in multiple tumor types. This research utilized bioinformatics coupled with pharmacological network approaches to explore the therapeutic significance alongside the underlying molecular pathways of MAT in BC. Differentially expressed genes (DEGs) associated with BC were retrieved from the gene expression omnibus (GEO) repository. miRNA targets were forecasted through the comparative toxicogenomics database (CTD) And SwissTargetPrediction. The overlap revealed potential therapeutic candidates, which were subsequently subjected to gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG) pathway enrichment, network analysis using cytoscape, machine learning approaches including least absolute shrinkage and selection operator (LASSO) and random forest (RF), as well as molecular docking studies. From the GEO database, a total of 5263 DEGs associated with BC were retrieved. Concurrently, 212 targets related to MAT were predicted utilizing CTD and SwissTargetPrediction. Cross-referencing these sets revealed 62 candidate therapeutic genes. Functional enrichment analyses including GO and KEGG suggested that these genes played key roles in apoptosis and cancer-related signaling cascades. A protein interaction network built in cytoscape highlighted 17 interconnected genes. By integrating LASSO and RF analyses, three central hub genes–cyclin D1 (CCND1), fibronectin 1 (FN1), and matrix metallopeptidase 13 (MMP13)-were identified. Molecular docking indicated strong affinity between MAT and these proteins, supporting their potential as critical targets for BC therapy. • 62 Genes were selected, showing enrichment in apoptosis and cancer-related pathways via GO and KEGG analyses. • Three hub genes (CCND1, FN1, and MMP13) were selected using LASSO and random forest algorithms. • Docking analysis demonstrated that matrine exhibited high affinity with the selected hub genes.