Assessing an ensemble docking-based virtual screening strategy for kinase targets by considering protein flexibility.

Area of Science:

• Computational chemistry
• Drug discovery
• Structural bioinformatics

Background:

• Receptor flexibility is crucial for accurate molecular docking.
• Traditional docking methods often use single, rigid receptor structures, limiting their effectiveness.
• Virtual screening (VS) requires robust protocols to identify potential drug candidates.

Purpose of the Study:

• To develop and evaluate a novel ensemble docking protocol.
• To improve the prediction accuracy of docking-based virtual screening.
• To accommodate receptor flexibility in drug discovery.

Main Methods:

• Developed an ensemble docking protocol using naïve Bayesian classification.
• Utilized structural clustering to select representative conformations from crystal structures and molecular dynamics (MD) simulations.
• Evaluated the protocol's performance on kinase targets: ALK, CDK2, and VEGFR2.

Main Results:

• Ensemble docking integrating multiple conformations outperformed single-structure docking.
• Crystal structures, on average, showed better discrimination of inhibitors than MD-derived structures.
• The integrated VS strategy demonstrated superior performance in virtual screening accuracy.

Conclusions:

• The novel ensemble docking protocol offers improved virtual screening capabilities.
• This method enhances the search for diverse and promising active compounds.
• The protocol provides a valuable advancement over existing virtual screening strategies.