Efficiently Transplanting Potential Energy Interpolation Database between Two Systems: Bacteriochlorophyll Case with...

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Biophysics

Background:

• Accurate potential energy surfaces (PES) are crucial for computational simulations of molecular dynamics.
• Interpolation mechanics/molecular mechanics (IM/MM) offers quantum chemistry (QC) accuracy at a reduced computational cost, but database construction is challenging.

Purpose of the Study:

• To develop a method for constructing IM databases for one system using preexisting data from a related system.
• To demonstrate this method by creating a PES for bacteriochlorophylls (BChls) in the light-harvesting 2 (LH2) complex.

Main Methods:

• Transplanting an existing IM database from the Fenna-Matthews-Olson (FMO) complex to the LH2 complex.
• Utilizing BChl displacement vectors to account for protein scaffold geometry differences.
• Employing a genetic algorithm for selecting effective data points from the transplanted database.
• Using two levels of basis sets for QC calculations to expedite the selection process.
• Performing 1 ns IM/MM molecular dynamics simulations to validate the constructed PES.

Main Results:

• The developed method successfully constructed an IM database for BChls in the LH2 complex.
• The resulting PES showed high accuracy, with small errors in ground-state, excited-state, and transition energies (∼0.07 eV and ∼0.008 eV, respectively).
• The IM/MM simulations confirmed the reliability of the generated PES.

Conclusions:

• The presented method provides an efficient way to construct IM databases by leveraging existing data from related systems.
• Effective data point selection is critical for generating accurate IM PES for molecular dynamics simulations.
• This approach significantly reduces the challenges associated with building high-accuracy PES for complex biological systems.