<p>Breast cancer is a malignant tumor originating from the breast epithelium, and emerging evidence suggests that the gut microbiota influences its development, progression, and treatment, although its role remains underexplored. In this study, we employed an integrative multi-omics framework that combined network pharmacology, machine learning, SHapley Additive exPlanations (SHAP), and single-cell RNA sequencing to systematically investigate key interactions between microbial metabolites and their targets. Core regulators were further validated using Mendelian randomization (MR), while molecular docking was applied to evaluate the binding affinity of candidate metabolites. Matrix metalloproteinase-3 (MMP3) emerged as a central molecule involved in multiple cancer-related signaling pathways, including PI3K-AKT, MAPK, and HIF-1, with promising druggable potential. Eight non-toxic gut microbial metabolites—such as indole-3-propionic acid, glycocholic acid, and 4-hydroxyphenylpyruvate—demonstrated strong binding affinity to MMP3 and favorable pharmacokinetic properties, highlighting a previously unappreciated microbiota–MMP3 axis as a promising avenue for therapeutic intervention in breast cancer. These findings provide a basis for subsequent in vitro and in vivo validation and underscore the translational potential of the identified microbial metabolites, thereby supporting the development of microbiome-derived therapeutic strategies for breast cancer.</p>