Energy-Screened Many-Body Expansion for Protein-Ligand Interactions: Examining Convergence for Metalloenzymes Through...

Area of Science:

• Computational chemistry
• Biochemistry
• Quantum mechanics

Background:

• Fragment-based quantum chemistry is crucial for protein-ligand interaction energy calculations.
• Traditional methods require large models and face computational challenges with many-body expansions.
• Instability arises from combinatorial complexity in subsystem calculations.

Purpose of the Study:

• To develop an energy-based screening protocol for the many-body expansion (MBE) in protein-ligand interactions.
• To implement this protocol in the open-source Fragme∩t code.
• To improve the efficiency and stability of quantum chemistry calculations for biomolecular systems.

Main Methods:

• Utilized aggressive screening based on semiempirical quantum chemistry.
• Employed an improved graph-theoretical algorithm to identify and eliminate unimportant subsystems.
• Performed n-body calculations up to n=7 using density functional theory with triple-ζ basis sets.
• Incorporated distance cutoffs to further optimize computational cost without sacrificing accuracy.

Main Results:

• Achieved rapid and stable convergence of the MBE by n=4 for metalloenzymes.
• Reduced or eliminated oscillations in n-body corrections compared to distance-cutoff-only methods.
• Demonstrated benchmark-quality protein-ligand interaction energies.
• Confirmed excellent parallel efficiency and scalability of the developed method.

Conclusions:

• The energy-based screening protocol offers a robust and efficient approach for calculating protein-ligand interaction energies.
• This method enables systematic convergence to benchmark quality.
• The Fragme∩t code implementation provides a scalable solution for complex biomolecular systems.