Abstract
Dehydrosqualene synthase is a validated antibacterial target involved in Staphylococcus aureus staphyloxanthin biosynthesis, a key factor in oxidative stress resistance and virulence. From a dataset of 299,845 natural products, we applied pharmacophore-based screening followed by QSAR-based screening, identifying 1,978 compounds with predicted Ki values of 100 nM or lower. Among them, 47 molecules showed remarkable predicted activity, with Ki values lower than 1.0 nanomolar. Notably, despite the wide chemical diversity of the library evaluated and the fact that the models were not trained with certain scaffolds of known inhibitors of staphyloxanthin biosynthesis, nine of these chemotypes were recovered among these top candidates, demonstrating the ability of the models to capture relevant structural features beyond the training data. Five representative compounds from this group showed high binding affinity and stable interactions with key CrtM residues in molecular dynamics simulations, reflecting the potential of the overall set as a source of promising leads for novel antivirulence agents targeting dehydrosqualene synthase in S. aureus.