Refinement and Rescoring of Virtual Screening Results

Area of Science:

• Computational chemistry
• Drug discovery
• Molecular modeling

Background:

• High-throughput docking is a key computational method for identifying potential drug candidates.
• Current docking methods face limitations in scoring functions and sampling accuracy, leading to approximate results.
• Post-docking analysis is crucial for validating and improving docking outcomes.

Purpose of the Study:

• To present BEAR (Binding Estimation After Refinement), a novel post-docking analysis tool.
• To enhance the accuracy of virtual screening by refining docking poses and employing advanced rescoring techniques.
• To discuss future directions in computational drug discovery refinement and rescoring.

Main Methods:

• Development of the BEAR tool for post-docking analysis.
• Utilizing molecular dynamics (MD) simulations to refine ligand-target complex poses.
• Applying MM-PB(GB)SA scoring functions for more accurate ligand rescoring.

Main Results:

• BEAR refines docking poses and rescores ligands for improved accuracy.
• The tool has been successfully validated in computational drug discovery.
• Prospective applications in drug discovery demonstrate the tool's utility.

Conclusions:

• BEAR offers a robust approach to post-docking analysis, enhancing virtual screening reliability.
• Refinement and rescoring are essential for overcoming inherent limitations of docking methodologies.
• The developed methodology shows promise for future drug discovery endeavors.