Specific non-local interactions are not necessary for recovering native protein dynamics

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• The elastic network model (ENM) is a standard method for analyzing protein dynamics using normal mode analysis (NMA).
• ENM relies on O(N^2) pairwise distance information for proteins with N residues, which is computationally intensive.
• Understanding the minimum information required for accurate protein dynamics simulation is crucial for computational efficiency.

Purpose of the Study:

• To develop and evaluate a novel model for protein dynamics that requires significantly less structural information than ENM.
• To compare the performance of the new model against ENM and explicit-solvent molecular dynamics (MD) simulations.
• To assess the ability of the new model to reproduce experimental B-factors.

Main Methods:

• Introduction of the 'contact number diffusion' (CND) model, utilizing O(N) restraints.
• CND employs local atom pair restraints and semi-specific non-local restraints based on atomic contact numbers.
• Comparison of CND dynamics with ENM and explicit-solvent MD simulations, and experimental B-factors.

Main Results:

• The CND model reproduces protein dynamics comparable to ENM but shows higher correlation with explicit-solvent MD simulations.
• CND effectively suppresses unrealistic surface fluctuations commonly seen in ENM.
• While ENM may capture some specific non-local interactions better, CND's semi-specific restraints are largely sufficient for native protein dynamics.

Conclusions:

• The contact number diffusion (CND) model offers a computationally efficient alternative to ENM for studying protein dynamics.
• CND provides a more accurate representation of protein dynamics compared to ENM, particularly in its correlation with MD simulations.
• The findings suggest that a reduced set of restraints, focusing on local and semi-specific non-local interactions, can effectively capture essential protein dynamics.