Efficient solvent boundary potential for hybrid potential simulations.

Area of Science:

• Computational biophysics
• Quantum chemistry
• Molecular mechanics

Background:

• Simulating large biological systems computationally is challenging due to finite system sizes.
• Accurate modeling of catalytic reactions requires efficient methods that capture both quantum and environmental effects.

Purpose of the Study:

• To develop and implement a computationally efficient hybrid quantum chemical/molecular mechanical (QC/MM) algorithm with a solvent boundary potential.
• To accurately model statistical properties of infinite bulk systems from finite simulations.

Main Methods:

• A hybrid QC/MM approach partitioning the system into quantum, explicit solvent (MM), and implicit solvent regions.
• A solvent boundary potential is constructed to mimic bulk solvent effects, including electrostatics and van der Waals interactions.
• The method was implemented in the pDynamo simulation program.

Main Results:

• The algorithm was tested on enzyme reaction mechanisms (citrate synthase, lactate dehydrogenase).
• Calculated energies and geometries of reaction intermediates showed good agreement with experimental and previous computational studies.
• The method effectively reproduces solvation free energy for embedded atoms.

Conclusions:

• The developed QC/MM method with a solvent boundary potential is computationally efficient and accurate for studying enzyme catalysis.
• This approach provides a reliable way to obtain bulk system properties from finite simulations.
• Further improvements to the method's utility are possible and discussed.