Dynamic Docking Using Multicanonical Molecular Dynamics: Simulating Complex Formation at the Atomistic Level.

Area of Science:

• Computational chemistry
• Molecular dynamics
• Structural biology

Background:

• Predicting protein-ligand binding configurations is crucial for drug discovery.
• Traditional molecular dynamics (MD) simulations face challenges in adequately sampling conformational and configurational space.
• Enhanced sampling techniques are needed to explore diverse binding modes.

Purpose of the Study:

• To present a multicanonical molecular dynamics (McMD)-based dynamic docking methodology.
• To demonstrate the capability of McMD to predict native binding configurations for protein-ligand systems.
• To showcase an efficient approach for sampling complex conformational and configurational spaces.

Main Methods:

• Application of multicanonical molecular dynamics (McMD) for enhanced sampling.
• Dynamic docking to predict binding configurations.
• Principal component analysis (PCA) to analyze the sampled ensemble.
• Free energy landscape (FEL) construction and analysis.
• Ranking of representative structures based on free energy.

Main Results:

• McMD successfully sampled bound and unbound ligand configurations and receptor conformations.
• The methodology enabled efficient exploration of conformational and configurational space beyond canonical MD.
• Analysis of the reweighted ensemble via PCA and FEL revealed key binding landscapes.
• Representative structures at local FEL minima were identified and ranked.
• The dynamic docking method accurately reproduced native binding configurations for various ligand types.

Conclusions:

• McMD-based dynamic docking is a powerful technique for predicting native protein-ligand binding modes.
• This approach overcomes sampling limitations of conventional MD simulations.
• The methodology is effective for small compounds, medium-sized compounds, and peptides.
• This work provides a robust framework for computational drug design and discovery.