Scalable Indirect Free Energy Method Applied to Divalent Cation-Metalloprotein Binding

Area of Science:

• Computational chemistry and biophysics
• Molecular modeling and simulation

Background:

• Accurate molecular simulations of biological systems with charged molecules require advanced electrostatic models.
• Fixed partial charge models lack experimental accuracy, while polarizable models (e.g., AMOEBA) are computationally expensive.
• Existing indirect free energy (IFE) methods struggle to reweight more than ~50 atoms due to scaling issues.

Purpose of the Study:

• To develop a novel indirect free energy method capable of reweighting thousands of atoms for molecular simulations.
• To improve the computational efficiency and accuracy of modeling electrostatic interactions in large biomolecular systems.

Main Methods:

• Introduction of "simultaneous bookending" (SB), a tunable sampling approximation for IFE methods.
• Application of SB to calculate the relative binding affinity of Mg2+/Ca2+ to metalloproteins (up to 2972 atoms).
• Comparison of SB results with direct polarizable AMOEBA simulations.

Main Results:

• SB successfully reweighted systems with thousands of atoms, overcoming previous limitations.
• No statistically significant difference was observed between direct AMOEBA and Amber-to-AMOEBA corrected results using SB.
• The method demonstrated scalability for large biomolecular systems.

Conclusions:

• Simultaneous bookending (SB) offers a computationally efficient and accurate approach for molecular simulations involving large numbers of atoms.
• SB significantly expands the applicability of advanced electrostatic models to complex biological problems.
• Future applications may include protein-protein binding and nucleic acid thermodynamics.