A Multiscale Simulation Approach to Compute Protein-Ligand Association Rate Constants by Combining Brownian Dynamics...

Area of Science:

• Computational chemistry
• Biophysics
• Pharmacology

Background:

• Experimental determination of drug-protein binding kinetics is costly and time-consuming.
• Accurate computational prediction requires balancing simulation fidelity with computational feasibility.
• Existing Brownian dynamics (BD) and molecular dynamics (MD) combinations often demand significant computational resources.

Purpose of the Study:

• To develop a computationally efficient multiscale pipeline for predicting drug-protein binding kinetics.
• To optimize the combination of BD and MD simulations for improved accuracy and speed.
• To validate the pipeline across diverse protein-ligand systems.

Main Methods:

• Developed a multiscale computational pipeline integrating BD and MD simulations.
• Optimized BD sampling to generate an ensemble of diffusional encounter complexes.
• Utilized these complexes as starting points for MD simulations to capture binding events.
• Validated the pipeline on various protein-ligand complexes with different properties.

Main Results:

• The multiscale pipeline significantly improves computational efficiency compared to existing methods.
• Predicted association rate constants (k_on) show good agreement with experimental measurements.
• The approach provides insights into the physical factors governing binding rates.
• Successful validation across diverse protein-ligand systems demonstrates broad applicability.

Conclusions:

• The developed multiscale BD/MD pipeline offers a computationally efficient and accurate method for determining drug-protein binding kinetics.
• This approach reduces the cost and time associated with kinetic parameter estimation.
• The validated pipeline can aid in drug discovery and development by providing reliable kinetic insights.