The Science and Art of Structure-Based Virtual Screening

Area of Science:

• Computational chemistry and cheminformatics
• Drug discovery and development
• Medicinal chemistry

Background:

• High attrition rates in drug discovery necessitate exploring larger chemical spaces.
• Structure-based virtual screening (SBVS) is regaining prominence as a key strategy.
• The Bayesian perspective highlights limitations in sustainable hit discovery.

Purpose of the Study:

• To discuss the shortcomings of SBVS as a sustainable hit discovery strategy.
• To analyze the impact of prior hit rates and computational method performance on SBVS.
• To share practical lessons learned from successful SBVS campaigns for experimental validation.

Main Methods:

• Review of Bayesian perspectives on virtual screening.
• Analysis of prior hit rates in screening libraries.
• Evaluation of computational method performance in SBVS.
• Case studies from eight successful SBVS campaigns.

Main Results:

• Shortcomings of SBVS identified from a Bayesian viewpoint.
• Prior hit rates and computational performance critically influence SBVS success.
• Eight campaigns yielded valuable insights into virtual hit selection.
• One campaign contributed to a drug candidate now in clinical trials.

Conclusions:

• SBVS requires careful consideration of library quality and computational tools for sustainable hit discovery.
• Lessons from practical campaigns enhance the selection of virtual hits for validation.
• Optimized SBVS strategies can significantly impact drug candidate progression.