EPB

EPB incorporates polarization into electrostatic interaction calculations to improve protein–ligand docking accuracy and binding energy estimation.

Key Features:

• Polarizable bond model: Introduces a polarizable bond representation specifically tailored for small organic molecules.
• Accounting for polarization: Replaces fixed point charge–charge approximations to capture dynamic polarization effects in biomolecular environments.
• Polarized ligand charges: Computes polarized ligand charges from protein–ligand complex structures for use in interaction energy calculations.
• Integration with docking (EPB Dock): Integrates EPB charges into protein–ligand docking workflows (EPB Dock) to influence pose ranking and scoring.
• Improved interaction terms: Provides more accurate representations of van der Waals and electrostatic interactions, including intermolecular hydrogen bonding.
• Benchmarking on PDB complexes: Applied to 38 cocrystallized structures from the Protein Data Bank (PDB) for method evaluation.
• Quantitative performance metrics: Demonstrates reductions in maximum error (from 7.98 Å to 2.03 Å) and case-specific improvements such as for PDB 1fqx (maximum error 12.88 Å to 1.57 Å and average RMSD 2.83 Å to 1.85 Å).

Scientific Applications:

• Protein–ligand docking: Improves pose selection and scoring by incorporating polarization-sensitive electrostatics.
• Binding energy estimation: Enhances estimation of ligand–protein binding energies through polarized charge calculations.
• Hydrogen-bond modeling: Improves representation of intermolecular hydrogen bonding in docking and interaction analyses.
• Benchmarking and validation: Enables evaluation of docking protocols using cocrystallized PDB structures, including case studies such as PDB 1fqx.

Methodology:

Introduces a polarizable bond model for small organic molecules, computes polarized ligand charges from protein–ligand complex structures, integrates these charges into docking (EPB Dock), and benchmarks performance on 38 cocrystallized PDB structures using maximum error and RMSD metrics.