Multi-conformer ensemble docking to difficult protein targets.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Traditional molecular docking often struggles with protein targets lacking high-resolution crystal structures.
• The Directory of Useful Decoys (DUD) provides a benchmark for evaluating docking performance.
• Protein flexibility is a key challenge in accurate molecular docking.

Purpose of the Study:

• To assess the efficacy of large-scale ensemble docking using molecular dynamics (MD) simulations.
• To improve hit enrichment and chemical diversity in virtual screening against difficult protein targets.
• To compare ensemble docking performance against crystal structure docking and random selection.

Main Methods:

• Generated molecular dynamics trajectories for five challenging protein targets from the DUD database.
• Extracted representative conformational structures from MD trajectories to form docking ensembles.
• Performed ensemble-based docking calculations and compared enrichment factors with crystal structure docking.

Main Results:

• Ensemble docking using simulation-derived structures yielded improved enrichment factors for four out of five proteins.
• The simulation-based approach increased the chemical diversity of identified docking hits.
• Docking to crystal structures or random compound selection showed lower performance.

Conclusions:

• Molecular dynamics-derived ensembles offer a superior strategy for virtual screening of difficult protein targets.
• Combining MD simulations with ensemble docking enhances hit identification and diversity.
• This approach is effective for large-scale chemical library screening in drug discovery.