Developing a molecular dynamics force field for both folded and disordered protein states

Abstract

SignificanceMany proteins that perform important biological functions are completely or partially disordered under physiological conditions. Molecular dynamics simulations could be a powerful tool for the structural characterization of such proteins, but it has been unclear whether the physical models (force fields) used in simulations are sufficiently accurate. Here, we systematically compare the accuracy of a number of different force fields in simulations of both ordered and disordered proteins, finding that each force field has strengths and limitations. We then describe a force field that substantially improves on the state-of-the-art accuracy for simulations of disordered proteins without sacrificing accuracy for folded proteins, thus broadening the range of biological systems amenable to molecular dynamics simulations.

Temperature, Hydrostatic Pressure, and Osmolyte Effects on Liquid–Liquid Phase Separation in Protein Condensates: Physical Chemistry and Biological Implications

“…[125] Attempts have been made to improve the performance of forcefields in IDP applications, with the introductiono fm odified forcefields a03ws, [126] CHARMM36m, [127] and a992B-disp. [128] But none of the modified forcefields proposed thus far is accurate for both globular proteins and IDPs, raising the question as to whether the mathematical functional forms of the pairwise interactions in current forcefield are fundamentally limiting and that context dependenceo rm ultiple-body effectsh ave to be incorporated to capture biomolecular behaviors more accurately. [129][130][131] In this context, progress has nonetheless been made in theoreticalp hysical understanding of how LLPS of IPDs might depend on their amino acid sequences by using coarse-grained approaches.…”

Perspective: Chain dynamics of unfolded and intrinsically disordered proteins from nanosecond fluorescence correlation spectroscopy combined with single-molecule FRET

“…Triggered primarily by the increasing interest in IDPs, recent developments have changed this situation considerably. 7,11,12,[102][103][104][105][106] Another important development has been the use of experimental restraints, where a simulated conformational ensemble is biased or reweighted based on experimental information, taking into account the uncertainties of the measurements. 85,[107][108][109] These developments have started to enable a more and more realistic and detailed view of unfolded proteins.…”

Force‐field parametrization based on radial and energy distribution functions

“…This newly developed model was named SPC/E b . In the TIP4P/2005 water model, an increase in the dispersion term followed by adjusting the repulsive term improved the structure reproducibility of disordered proteins, which was named as TIP4P‐D and was revised recently …”

“…[8] In the TIP4P/2005 water model, [9] an increase in the dispersion term followed by adjusting the repulsive term improved the structure reproducibility of disordered proteins, [10] which was named as TIP4P-D and was revised recently. [11] In this study, we propose an approach that can automatically determine the parameters of a molecule. A molecule of interest is described with energy functions of bonded and nonbonded parameters in an additive fashion as done in common FFs such as AMBER FFs.…”