Exploring hierarchical refinement techniques for induced fit docking with protein and ligand flexibility

Area of Science:

• Computational Biology
• Structural Biology
• Biochemistry

Background:

• Accurate prediction of protein-ligand complex structures is crucial for drug discovery and understanding biological processes.
• Induced fit docking methods aim to model conformational changes in proteins upon ligand binding.
• Existing methods often struggle to balance ligand flexibility with necessary protein backbone rearrangements.

Purpose of the Study:

• To present novel molecular mechanics-based protein refinement methods for induced fit docking.
• To evaluate the performance of these methods on a diverse set of protein-ligand complexes.
• To identify optimal strategies for refining protein-ligand interactions, emphasizing active site dynamics.

Main Methods:

• Development and application of two novel molecular mechanics-based refinement techniques.
• Integration of minimization, sidechain prediction, hierarchical ligand placement, and minimized Monte Carlo with normal mode analysis.
• Testing on 88 protein-ligand complexes, including cross-docking and apo-docking scenarios.

Main Results:

• Prioritizing active site backbone opening is essential and can be hindered by focusing solely on ligand flexibility.
• A minimized Monte Carlo procedure for active site opening, followed by hierarchical sidechain packing optimization around a flexible ligand, yielded the highest accuracy.
• Achieved within 2.0 Å heavy-atom RMSD for 75% of complexes when protein backbone rearrangement was < 1.0 Å alpha-carbon RMSD.
• Demonstrated that physics-based all-atom potentials outperform docking potentials with sufficient refinement.

Conclusions:

• The presented refinement methods significantly improve the accuracy of induced fit docking.
• Optimizing protein backbone flexibility, particularly active site opening, is critical for successful protein-ligand complex modeling.
• The minimized Monte Carlo approach combined with hierarchical sidechain packing represents a powerful strategy for accurate structural prediction.