Improving VEGFR-2 docking-based screening by pharmacophore postfiltering and similarity search postprocessing

Area of Science:

• Computational chemistry
• Medicinal chemistry
• Drug discovery

Background:

• Virtual screening (VS) relies on scoring functions to rank compounds, but their effectiveness varies with receptor pockets.
• Distinguishing correct binding modes from incorrect ones remains a challenge in conventional VS.

Purpose of the Study:

• To develop and validate a novel three-step VS protocol to enhance the selection of active compounds.
• To improve both overall and early virtual screening performance.

Main Methods:

• A three-step VS protocol integrating docking, pharmacophore postfiltering, and similarity search postprocessing.
• Retrospective application to VEGFR-2 (Kdr-kinase) inhibitors using MOE's Alpha HB scoring function.
• Postfiltering based on structure-based pharmacophore interactions and re-ranking using MACCS fingerprints.

Main Results:

• The proposed three-step VS protocol significantly improves virtual screening performance.
• Both overall and early performance metrics showed enhancement compared to conventional methods.
• The protocol effectively distinguishes active compounds from decoys.

Conclusions:

• The integrated VS protocol offers a more effective approach for identifying potential drug candidates.
• Pharmacophore filtering and similarity searching enhance the accuracy of docking-based VS.
• This method provides a valuable tool for accelerating drug discovery efforts.