Energy Decomposition Analysis of Protein-Ligand Interactions Using Molecules-in-Molecules Fragmentation-Based Method

Area of Science:

• Computational Chemistry
• Structural Biology
• Drug Discovery

Background:

• Predicting protein-ligand binding affinities is crucial but computationally expensive.
• Quantum mechanical (QM) methods offer accuracy but are often too slow for broad application.
• Previous work showed success with the molecules-in-molecules (MIM) fragmentation method for gas-phase interactions.

Purpose of the Study:

• To introduce a novel three-layer model (MIM3) for calculating protein-ligand interactions and binding affinities.
• To decompose binding affinity into contributions from residues, water, solvation, and entropy.
• To enable accurate and cost-effective analysis for drug design.

Main Methods:

• Utilized a three-layer model (MIM3) combining QM for intermolecular interactions with empirical models for desolvation and entropy.
• Applied the MIM3 approach to a series of 22 thrombin inhibitors with known binding affinities.
• Calculated binding affinities using B97-D3BJ/6-311++G(2d,2p) QM level on a 120-residue binding pocket.

Main Results:

• MIM3 showed good correlation with experimental binding affinities (R²: 0.81-0.88, ρ: 0.84-0.89).
• The method provided quantitative, residue-specific interaction energies.
• Identified both common (e.g., hydrogen bonds) and uncommon (e.g., CH···π) interactions crucial for binding.

Conclusions:

• The MIM3 approach offers an accurate and computationally efficient way to predict protein-ligand binding affinities.
• Residue-specific interaction data can guide the optimization of drug candidates.
• This method can improve the effectiveness of computational drug design by identifying key binding determinants.