Sensitivity of an empirical affinity scoring function to changes in receptor-ligand complex conformations.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Accurate prediction of ligand binding affinity is crucial for drug discovery.
• Scoring functions are widely used in computational drug design.
• Conformational flexibility of ligands and proteins can impact binding affinity.

Purpose of the Study:

• To investigate the sensitivity of empirical scoring functions to conformational changes in ligands.
• To evaluate the performance of scoring functions using molecular dynamics simulations.
• To understand how different binding site properties influence scoring function accuracy.

Main Methods:

• Utilized empirical scoring functions combined with conformational sampling.
• Employed molecular dynamics simulations to generate ensembles of protein-ligand complex structures.
• Analyzed the correlation between calculated scores and structural deviations from experimental data.

Main Results:

• Scoring function sensitivity to ligand conformation varied depending on the binding site.
• Clear correlation observed for hydrogen-bond dominated complexes (arabinose-binding protein).
• Hydrophobic complexes (retinol-binding protein) showed less sensitivity to ligand conformational changes.
• Score variations from molecular dynamics ensembles were smaller than molecular mechanics interaction energy variations.

Conclusions:

• The effectiveness of scoring functions in predicting binding affinity is influenced by ligand conformation and binding site characteristics.
• Molecular dynamics simulations provide valuable structural ensembles for assessing scoring function robustness.
• Further refinement of scoring functions may be needed to account for diverse binding modes.