Published March 4, 2023
| Version v1
Preprint
Open
Structure-Based Experimental Datasets for Benchmarking of Protein Simulation Force Fields
Authors/Creators
- 1. UC San Diego
- 2. Rutgers U
- 3. Los Alamos National Laboratory
- 4. U. of Pittsburgh
- 5. City College of New York
- 6. University of Copenhagen
- 7. UC Irvine
- 8. U. of Helsinki
- 9. U of Natural Resources and Life Sciences, Vienna
- 10. Dartmouth College
- 11. Temple University
Description
This review article provides an overview of structurally oriented, experimental datasets that can be used to benchmark protein force fields, focusing on data generated by nuclear magnetic resonance (NMR) spectroscopy and room temperature (RT) protein crystallography. We discuss why these observables are useful for assessing force field accuracy, how they can be calculated from simulation trajectories, and statistical issues that arise when comparing simulations with experiment. The target audience for this article is computational researchers and trainees who develop, benchmark, or use protein force fields for molecular simulations.
Files
V5 Protein benchmark data draft for Zenodo-3-3-2023.pdf
Additional details
Funding
- Research Council of Finland
- Combined approach to understand biological lipid-protein interfaces 319902
- Research Council of Finland
- Combined approach to understand biological lipid-protein interfaces 315596