Variational Optimization of an All-Atom Implicit Solvent Force Field to Match Explicit Solvent Simulation Data.

Area of Science:

• Computational Biophysics
• Molecular Modeling
• Protein Simulations

Background:

• Accurate implicit solvation models are crucial for large-scale biophysical simulations.
• Existing models often lack efficiency or accuracy for disordered protein states.
• Need for computationally inexpensive methods that overlap with explicit solvent simulations.

Purpose of the Study:

• To develop an efficient implicit solvation model (EEF1-SB) for protein simulations in aqueous environments.
• To ensure good overlap with explicit solvent simulations, especially for unfolded and disordered protein states.
• To enable multiscale applications requiring accurate treatment of solvent effects.

Main Methods:

• Developed an efficient solvation term based on a Gaussian solvent-exclusion model (EEF1).
• Employed a coarse-graining procedure minimizing an entropy-related objective function to train the model.
• Optimized charge screening and backbone torsion terms against explicit solvent simulations of peptides.

Main Results:

• The resulting effective energy function, EEF1-SB, shows reasonable performance across various systems.
• EEF1-SB excels at capturing the structure and dimensions of disordered or weakly structured peptides.
• EEF1-SB is approximately 10 times faster than Generalized-Born methods.

Conclusions:

• EEF1-SB provides a computationally inexpensive and transferable approximation for solvent effects in protein simulations.
• The model demonstrates particular strength in accurately simulating disordered protein states.
• EEF1-SB is well-suited for multiscale biophysical applications requiring efficient and accurate solvation treatment.