Protein Alpha Shape (PAS) Dock: a new gaussian-based score function suitable for docking in homology modelled protein...

Area of Science:

• Computational chemistry and structural biology.
• Drug discovery and development.

Background:

• Virtual screening is crucial for identifying potential drug candidates.
• Homology modeled protein structures present challenges for traditional docking methods due to inherent inaccuracies.
• Existing docking tools like AutoDock and MOE-Dock have limitations in computational efficiency and robustness against structural variations.

Purpose of the Study:

• To introduce and evaluate Protein Alpha Shape (PAS) Dock, a novel empirical score function for virtual screening.
• To assess the performance of PAS-Dock in predicting protein-ligand binding free energies, particularly with homology models.
• To compare PAS-Dock's accuracy and computational efficiency against established docking software.

Main Methods:

• PAS-Dock utilizes Gaussian property fields to describe protein binding sites and ligand properties, maximizing favorable interactions (hydrophilicity, lipophilicity) while minimizing steric clashes.
• A two-level scoring approach is employed: a simplified version for efficient geometry search (Tabu search) and a detailed version for binding free energy prediction.
• The score function was trained on 218 X-ray protein-ligand complexes with experimental binding affinities.

Main Results:

• PAS-Dock demonstrates superior computational efficiency compared to AutoDock and MOE-Dock.
• The method achieves better prediction accuracy for binding free energies.
• PAS-Dock exhibits greater robustness against small structural variations in protein models than AutoDock.

Conclusions:

• PAS-Dock is a computationally efficient and accurate tool for virtual screening, especially suitable for homology modeled protein structures.
• Its robustness and improved performance make it a valuable alternative for drug discovery pipelines.
• The Gaussian smoothing of property fields contributes to its resilience against structural variations and reduces the need for explicit protein flexibility.