Unveiling the full potential of flexible receptor docking using multiple crystallographic structures.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Receptor flexibility is a critical, yet challenging, aspect of molecular docking.
• Therapeutic target binding sites often adopt diverse conformational states impacting molecular recognition.
• Accurately modeling these dynamics is essential for effective drug design.

Purpose of the Study:

• To evaluate flexible receptor docking performance against rigid docking using extensive X-ray structures of CDK2 and HSP90.
• To assess improvements in binding-mode prediction and virtual screening efficacy.
• To identify limitations of flexible docking and propose solutions.

Main Methods:

• Utilized large datasets of X-ray crystallography structures for Cyclin Dependent Kinase 2 (CDK2) and Heat Shock Protein 90 (HSP90).
• Performed comparative analysis of flexible versus rigid receptor docking for binding-mode prediction.
• Conducted virtual screening experiments to evaluate binding affinity prediction and hit rates.
• Developed and tested a scoring scheme to mitigate false positives in flexible docking.

Main Results:

• Flexible receptor docking significantly outperforms rigid docking in predicting binding modes.
• Binding affinity predictions showed marked improvement with flexible docking.
• However, an increase in false positives limited the improvement in hit rates during virtual screening.
• A novel scoring scheme was introduced to address this limitation.

Conclusions:

• Flexible receptor docking is superior for binding-mode prediction in drug discovery.
• While improving affinity prediction, flexible docking requires careful handling to avoid false positives.
• Guidelines for mitigating inherent pitfalls of flexible receptor docking are provided for enhanced virtual screening accuracy.