Structure-Based Experimental Datasets for Benchmarking Protein Simulation Force Fields [Article v1.0]

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Protein force fields are essential for molecular dynamics (MD) simulations.
• Accurate force fields are crucial for reliable prediction of protein structure and dynamics.
• Benchmarking force fields requires high-quality experimental data.

Purpose of the Study:

• To review experimental datasets for benchmarking protein force fields.
• To explain how nuclear magnetic resonance (NMR) spectroscopy and room temperature (RT) protein crystallography data can assess force field accuracy.
• To guide computational researchers in using these datasets for molecular dynamics (MD) simulations.

Main Methods:

• Focus on structurally oriented experimental datasets.
• Utilize data from NMR spectroscopy.
• Incorporate data from RT protein crystallography.
• Connect experimental observables to MD simulations.

Main Results:

• NMR and RT crystallography provide valuable observables for structure and dynamics.
• These observables are critical for assessing protein force field accuracy.
• Best practices for setting up and analyzing benchmark simulations are discussed.

Conclusions:

• Experimental data, particularly from NMR and RT crystallography, are vital for validating protein force fields.
• Proper utilization of these datasets enhances the reliability of MD simulations.
• This review serves as a guide for researchers developing or using protein force fields and machine learning models.