Pocket v.2: further developments on receptor-based pharmacophore modeling.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Pharmacophore models are crucial for understanding protein-ligand interactions and enhancing drug binding affinity.
• Automated methods for deriving pharmacophore models can streamline the drug design process.

Purpose of the Study:

• To introduce Pocket v.2, a standalone program for automated pharmacophore model derivation from protein-ligand complex structures.
• To evaluate the performance of Pocket v.2 in reproducing known pharmacophore models and its ability to handle conformational variations.

Main Methods:

• Pocket v.2 was developed based on the Pocket module of the LigBuilder program.
• The program automatically derives pharmacophore models from protein-ligand complex structures without manual intervention.
• Key pharmacophore features are automatically reduced to a manageable set.

Main Results:

• Pocket v.2 successfully reproduced established pharmacophore models for cyclin-dependent kinase 2, HIV-1 protease, estrogen receptor, and 17beta-hydroxysteroid dehydrogenase.
• The program demonstrated robustness by tolerating minor protein conformational changes upon ligand binding, yielding consistent pharmacophore models.
• Similar pharmacophore models were generated for different proteins binding the same ligand, suggesting potential for protein classification.

Conclusions:

• Pocket v.2 provides an efficient and automated approach for deriving pharmacophore models from protein structures.
• The software's ability to generate consistent models despite minor protein flexibility and its potential for protein classification highlight its utility in drug discovery and structural biology research.