Structure-based CoMFA as a predictive model - CYP2C9 inhibitors as a test case.

Area of Science:

• Computational chemistry
• Molecular modeling
• Drug discovery

Background:

• Predicting small compound affinity to target proteins is crucial for efficient drug design.
• Accurate prediction models can significantly reduce the time and cost of identifying potential drug candidates.

Purpose of the Study:

• To establish a general computational scheme for predicting ligand-protein binding affinity.
• To develop a reliable method for generating accurate ligand-bound protein conformations.
• To construct a Quantitative Structure-Activity Relationship (QSAR) model for affinity prediction.

Main Methods:

• Identification of ligand-binding sites on protein targets.
• Generation of ligand conformations (poses) using molecular docking.
• Validation of correct ligand poses through consensus scoring and MM-PBSA analysis.
• Construction of a Comparative Molecular Field Analysis (CoMFA) model using validated poses.

Main Results:

• A CoMFA model was successfully developed using the crystal structure of CYP 2C9 and twenty known inhibitors.
• The model achieved good statistical performance, validating the accuracy of binding site classification and pose prediction.
• The developed scheme demonstrated reasonable accuracy in predicting ligand-protein binding affinities.

Conclusions:

• The proposed general scheme provides a robust method for predicting small compound affinity to target proteins.
• This approach enhances the reliability of computational chemistry in the drug design process.
• The validated methodology is expected to accelerate the discovery of novel therapeutics.