Assessment of Two Restraint Potentials for Coarse-Grained Chemical-Cross-Link-Assisted Modeling of Protein Structures

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Chemical cross-link mass spectrometry (XL-MS) provides distance restraints for protein structure modeling.
• Coarse-grained molecular dynamics (MD) simulations, like those using the UNRES force field, are valuable for modeling protein structures.
• Integrating experimental data into computational models is crucial for improving accuracy.

Purpose of the Study:

• To evaluate the impact of distance restraints from XL-MS on protein structure modeling quality using the UNRES force field.
• To compare the effectiveness of different types of restraining potentials for XL-MS data.

Main Methods:

• Utilized multiplexed replica exchange molecular dynamics simulations for 23 small proteins.
• Applied six types of cross-links with varying upper distance boundaries.
• Tested two restraining potentials: flat-bottom Lorentz-like and MD-based distance- and orientation-dependent potentials.

Main Results:

• Flat-bottom Lorentz-like potentials produced a higher number of higher-quality models compared to unrestrained simulations.
• MD-based potentials sometimes forced excessively long distances between side chains, reducing model quality.
• Significant model quality improvement was observed when the sum of differences in indices of cross-linked residues exceeded 150.

Conclusions:

• Recommends flat-bottom Lorentz-like potentials for representing cross-link restraints in protein structure modeling.
• Highlights the importance of proper parameterization for Lorentz-like potentials.
• Suggests a threshold for cross-linked residue index differences to ensure effective quality improvement.