Negative Image-Based Rescoring: Using Cavity Information to Improve Docking Screening.

Area of Science:

• Computational chemistry
• Drug discovery
• Molecular modeling

Background:

• Standard molecular docking struggles to accurately predict binding affinity, hindering the identification of effective drug candidates.
• Inaccurate estimation of binding free energy by default scoring functions leads to poor ranking of ligand poses and difficulty distinguishing active from inactive compounds.

Purpose of the Study:

• To introduce and describe the Negative image-based rescoring (R-NiB) method for enhancing virtual screening.
• To provide a practical guide for implementing R-NiB for improved drug discovery workflows.

Main Methods:

• The R-NiB method re-ranks molecular docking poses based on shape and electrostatic similarity between the ligand and the protein's binding cavity.
• The methodology has been validated across multiple drug targets and docking algorithms using benchmark datasets.

Main Results:

• R-NiB significantly improves the accuracy of virtual screening by effectively re-ranking docking poses.
• The method enhances the ability to differentiate between active and inactive compounds, increasing the yield of hit compounds.

Conclusions:

• Negative image-based rescoring (R-NiB) offers an effective and efficient solution to improve the outcomes of virtual screening assays.
• The R-NiB method's reliance on ligand-cavity complementarity makes it a robust tool for identifying novel drug candidates.