Glide WS: Methodology and Initial Assessment of Performance for Docking Accuracy and Virtual Screening

Area of Science:

• Computational chemistry
• Drug discovery
• Molecular modeling

Background:

• Growing interest in virtual screening for drug discovery.
• Need for performant methods to identify novel hits for challenging targets.
• Limitations of existing methods in pose prediction and virtual screening enrichment.

Purpose of the Study:

• Introduce Glide WS, a novel virtual screening method.
• Improve upon existing methods like Glide SP by incorporating explicit water representation.
• Enhance accuracy in pose prediction and virtual screening enrichment for drug discovery.

Main Methods:

• Developed Glide WS, an empirical scoring function for high-throughput docking.
• Incorporated explicit representation of water structure and dynamics.
• Tuned the scoring function using absolute binding free energy perturbation calculations (ABFEP).

Main Results:

• Glide WS achieved 92% self-docking accuracy, outperforming Glide SP (85%).
• Demonstrated significantly improved virtual screening enrichment across 38 diverse targets.
• Reduced the number of poorly scoring decoys compared to Glide SP, as validated by ABFEP.

Conclusions:

• Glide WS offers substantial gains in pose prediction and virtual screening enrichment.
• The explicit modeling of water improves the reliability of virtual screening for drug discovery.
• Glide WS is expected to yield higher hit rates in practical virtual screening applications compared to rigid receptor docking.