Ensemble-based docking: From hit discovery to metabolism and toxicity predictions

Area of Science:

• Computational chemistry
• Drug discovery
• Structural biology

Background:

• Traditional docking methods often use a single protein structure, which may not represent the dynamic nature of protein targets.
• This limitation can lead to missed opportunities in identifying potential drug candidates or inaccurate predictions.

Purpose of the Study:

• To describe and illustrate the application of ensemble-based docking for hit discovery.
• To explore its utility in understanding biochemical pathways and predicting drug candidate toxicity.
• To demonstrate its extension towards predicting metabolism and off-target binding.

Main Methods:

• Utilizing a collection of protein structures in docking calculations.
• Performing large-scale ensemble docking campaigns on supercomputers.
• Describing the necessary computational engineering for these campaigns.

Main Results:

• Ensemble-based docking significantly increased the number and diversity of validated drug candidates.
• Demonstrated successful application in hit discovery and lead optimization.
• Provided examples of its extension to toxicity and metabolism prediction.

Conclusions:

• Ensemble-based docking is a powerful strategy for enhancing drug discovery.
• It offers a structural basis for predicting drug metabolism and off-target interactions.
• This method is valuable for advancing drug candidates through preclinical and clinical development.