ColDock: Concentrated Ligand Docking with All-Atom Molecular Dynamics Simulation.

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Accurate prediction of protein-ligand complex structures is crucial for drug discovery.
• Traditional docking methods can struggle with spontaneous binding and accurate pose prediction.
• Molecular dynamics (MD) simulations offer a powerful approach but require efficient protocols.

Purpose of the Study:

• To develop a simple, efficient, and accurate method for generating protein-ligand complex structures.
• To validate the proposed method's ability to predict native ligand poses.
• To assess the method's performance across different starting conditions (holo vs. apo structures).

Main Methods:

• Concentrated ligand Docking (ColDock) employs multiple independent MD simulations.
• Ligands are initially placed at high concentration (~100 mM) around the protein.
• Spontaneous binding is induced, followed by clustering of ligand poses to identify dominant clusters.

Main Results:

• ColDock successfully generated accurate protein-ligand complex structures similar to crystal structures starting from holo proteins.
• The method reproduced correct ligand poses in three out of four cases starting from apo structures.
• Performance was dependent on the initial accessibility of the ligand binding pocket, succeeding with open pockets.

Conclusions:

• ColDock is an effective method for predicting protein-ligand complex structures, particularly when binding pockets are open.
• The approach leverages high ligand concentration in MD simulations to enhance binding probability and pose accuracy.
• ColDock can be readily implemented using standard MD simulation software for drug discovery applications.