Evaluation of ligand overlap by atomic parameters.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Ligand-protein interactions are crucial for drug development.
• Understanding atom-level interactions aids in predicting drug efficacy.
• 3D Quantitative Structure-Activity Relationship (3D QSAR) methods rely on physicochemical properties.

Purpose of the Study:

• To analyze ligand atom overlap in 32 common enzyme systems.
• To validate the use of atom-based physicochemical parameters in 3D QSAR.
• To introduce and assess a new statistic, mean maximal atomic deviation (MeMAD).

Main Methods:

• Superposition of experimentally determined protein structures to align ligands.
• Analysis of overlapping ligand atoms based on partition coefficient and molar refractivity.
• Calculation and validation of the mean maximal atomic deviation (MeMAD) against randomized data.

Main Results:

• Ligand atoms predominantly overlap with atoms of similar chemical nature.
• The newly developed MeMAD statistic was validated across all studied systems.
• MeMAD values were consistently distinct from randomized data, confirming their statistical significance.

Conclusions:

• Atom-based physicochemical parameters are reliable for 3D QSAR.
• The MeMAD statistic provides a robust measure for assessing ligand alignment validity.
• This study supports the foundation of structure-based drug design using physicochemical properties.