FWAVina

FWAVina optimizes protein-ligand docking by integrating the fireworks algorithm for global search and the Broyden-Fletcher-Goldfarb-Shannon (BFGS) local search method within the AutoDock Vina framework to identify optimal ligand poses in protein binding pockets.

Key Features:

• Protein-ligand docking optimization: Determines optimal ligand position and orientation within a target protein's binding pocket.
• AutoDock Vina framework: Builds upon the AutoDock Vina scoring and docking infrastructure.
• Fireworks algorithm (global optimization): Generates diverse candidate ligand poses through parallel global exploration.
• Broyden-Fletcher-Goldfarb-Shannon (BFGS) local search: Refines candidate poses via precise local optimization.
• Combined global-local pose search: Couples the fireworks global optimizer with BFGS local refinement to identify promising binding configurations.
• Benchmark evaluation: Performance was evaluated using the PDBbind and DUD-E datasets.
• Execution time improvement: Demonstrated execution time reductions of over 50% compared to the original Vina implementation.
• Prediction accuracy: Maintains high docking and virtual screening accuracy in comparative analyses.
• Robustness to ligand flexibility: Exhibits efficient performance with ligands having a high number of rotatable bonds.

Scientific Applications:

• Molecular docking: Identification and ranking of ligand binding poses within protein binding pockets.
• Virtual screening: High-throughput screening of compound libraries for potential binders using docking scores and poses.
• Computer-aided drug design: Support for the selection and optimization of candidate therapeutic compounds based on docking results.
• Benchmarking and method comparison: Comparative evaluation against AutoDock Vina using datasets such as PDBbind and DUD-E.

Methodology:

Implements a pose search that uses the fireworks algorithm for global optimization and the Broyden-Fletcher-Goldfarb-Shannon (BFGS) method for local refinement within the AutoDock Vina framework.