Integrated virtual and experimental screening of natural products as potential anticancer agents against breast cancer.

Breast cancer remains a leading cause of cancer-related mortality among women, driving the need for novel therapeutic agents. Here, we report an integrated in silico-in vitro screening pipeline to identify natural products with anticancer activity against key breast cancer targets. A library of 187,119 commercially available natural compounds from the ZINC database was subjected to molecular docking against five proteins implicated in breast tumorigenesis: mutant PIK3CA-E545K, overexpressed ESR1, mutant ERBB4-Y1242C, overexpressed EGFR, and overexpressed ERBB2. The top 15 compounds (C1-C15) were selected based on binding affinity and commercial availability, then evaluated for cytotoxicity in breast cancer cell lines (MCF-7, MDA-MB-468, SK-BR-3) and a normal fibroblast line (CCD-1064Sk). Several hits (most notably C3-C7 and C10) demonstrated potent binding (affinities ≤ -8.6 kcal/mol) with favorable selectivity indices (SI ≥ 2.0). A clear correlation between more negative docking scores and enhanced cytotoxicity supported the predictive value of our virtual screen. Structure-activity relationship analysis highlighted planarity and hydrophobic substituents as key drivers of activity. These findings validate our hybrid approach and identify promising natural-product leads for further preclinical evaluation in breast cancer models.