EADock: docking of small molecules into protein active sites with a multiobjective evolutionary optimization.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Protein-ligand docking is crucial for drug development, but high accuracy in predicting binding modes is needed for reliable free energy calculations and structure-based lead optimization.
• Existing high-throughput screening methods often lack the precision required for accurate binding mode identification.

Purpose of the Study:

• To present EADock, a novel docking software designed for high-accuracy protein-ligand binding mode prediction.
• To validate the efficacy of EADock's hybrid evolutionary algorithm and diversity management strategy.

Main Methods:

• EADock utilizes a hybrid evolutionary algorithm with dual fitness functions and sophisticated diversity management.
• The software is integrated with the CHARMM package for energy calculations and coordinate handling.
• Validation involved 37 crystallized protein-ligand complexes, defining the search space around crystal structures and using ligand positions with up to 10 Å RMSD.

Main Results:

• EADock successfully identified correct binding modes (RMSD < 2 Å) ranked first for 68% of complexes.
• Success rates improved to 78% within the top five ranked clusters and 92% considering all final generation clusters.
• Failures were often attributable to crystal contacts in experimental structures.

Conclusions:

• EADock demonstrates significant efficiency in sampling and accurately predicting protein-ligand binding modes.
• The software shows promise for structure-based drug optimization where experimental binding modes are unknown.
• Successful docking of RGD cyclic pentapeptide on alphaVbeta3 integrin validates EADock's real-world applicability.