Multi-Target Virtual Screening of Natural Products Against Virulence-Associated Proteins of Pseudomonas aeruginosa PAO1 for Antivirulence Candidate Prioritization

Background/Objectives: Natural products offer chemically diverse starting points for antivirulence discovery. We aimed to prioritize candidates against ten Pseudomonas aeruginosa PAO1 virulence-associated proteins while accounting for differences in structural support and compound-access requirements. Methods: We screened 391,658 COCONUT-derived ligand entries using staged AutoDock Vina docking, followed by GNINA rescoring and within-target Binding–Access–Exposure–Liability (BAEL) prioritization. Scaffold profiling, ranking sensitivity analyses, and triplicate 100 ns molecular dynamics simulations of 30 selected complexes characterized chemical diversity, selection consistency, and trajectory behavior. A separate retrospective LasB analysis examined sensitivity to comparison design. No prospective biochemical or cellular validation was performed. Results: The workflow yielded 189 candidate–target associations involving 182 unique natural products across six functional classes. The 181 scaffold-bearing compounds represented 176 Bemis–Murcko families. BAEL retained 179 of 200 Binding-only selections, with property-informed substitutions concentrated near the selection cutoffs. Top-20 selection overlap remained high under prespecified 25% perturbations of the Access, Exposure, and Liability rules (Jaccard similarity ≥ 0.905). Of 90 trajectories, 82 were classified as pocket-proximal under the modeled conditions. In the LasB analysis, areas under the receiver operating characteristic curve were 0.700/0.740 for Vina/GNINA with surrogate controls and 0.480/0.480 in the attempted property-matching sensitivity set, which showed residual property imbalance and differed in annotated-parent coverage and weighting. Conclusions: This study connects large-scale natural-product screening with target-specific experimental planning. Binding and antivirulence activity remain to be established, but the annotated candidates provide a chemically diverse starting point for focused biochemical and cellular evaluation.