Comparative analysis of various electrostatic potentials on docking precision against cyclin-dependent kinase 2...

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Drug Design

Background:

• Molecular modeling is crucial for understanding drug-receptor interactions.
• Accurate prediction of binding poses is essential for drug discovery.
• Various charge models exist, but their performance in docking studies varies.

Purpose of the Study:

• To assess the performance of different semiempirical and ab initio charge models in predicting docking poses against CDK2 proteins.
• To evaluate the effectiveness of multiple docking approaches and Prime/MM-GBSA calculations for predicting binding modes.
• To identify a reliable charge model and computational strategy for designing potent and selective CDK2 inhibitors.

Main Methods:

• Investigated semiempirical (RM1, AM1, PM3, MNDO) and ab initio (HF, DFT) charge models.
• Employed multiple docking approaches, including RRD (Rescoring with Rotamer Diversity) and IFD (Induced Fit Docking).
• Utilized Prime/MM-GBSA calculations to predict binding free energy and correlate with experimental activity.

Main Results:

• RRD demonstrated significant improvement in ligand binding poses and docking score accuracy compared to IFD.
• The combined RRD and Prime/MM-GBSA approach showed a high correlation between predicted binding free energy and experimental biological activity.
• The AM1 charge model emerged as a promising candidate for drug design, showing potential for enhanced success rates.

Conclusions:

• The AM1 charge model is a valuable tool for designing new drugs with improved success rates.
• The combination of RRD and Prime/MM-GBSA offers a robust method for predicting protein-ligand binding.
• The findings provide insights for developing potent and selective CDK2 inhibitors, applicable to other protein-ligand binding systems.