Random drift particle swarm optimisation algorithm for highly flexible protein-ligand docking.

Area of Science:

• Computational chemistry
• Drug discovery and development

Background:

• Molecular docking is crucial for drug design, but current methods struggle with flexible ligands.
• An increasing number of docking programs exist, yet limitations remain for complex molecular interactions.

Purpose of the Study:

• To introduce a novel hybrid algorithm, LRDPSO, for enhanced protein-ligand docking.
• To improve the accuracy of binding affinity prediction and ligand pose identification for flexible molecules.

Main Methods:

• Developed a hybrid algorithm combining Random Drift Particle Swarm Optimization and local search (LRDPSO) based on Autodock software.
• Focused on ligand flexibility and strategies to enhance binding affinity prediction in docking simulations.

Main Results:

• LRDPSO achieved significantly lower docking energy and higher precision than LGA, PSO, and QPSO algorithms.
• The LRDPSO algorithm correctly identified binding modes for 83.6% of complexes, outperforming QPSO (73.1%), PSO (68.7%), and LGA (68.7%).

Conclusions:

• The LRDPSO algorithm provides more reliable predictions for highly flexible ligands, especially in protein binding pockets where flexibility is key.
• This novel approach enhances the predictive power and broadens the applications of molecular docking in computational drug design.