Study of correlations between protein peptide plane dynamics and side chain dynamics

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Protein dynamics are essential for biological functions but challenging to study experimentally.
• All-atom molecular dynamics simulations have limitations in fully capturing protein dynamics.
• Reduced coordinate representations offer a promising alternative for analyzing protein dynamics.

Purpose of the Study:

• To investigate the predictability of side chain Cβ atom dynamics from static backbone (Cα and O) atom properties.
• To develop and validate a method for reconstructing all-atom protein dynamics from coarse-grained representations.
• To compare the precision of the developed method with existing all-atom reconstruction techniques.

Main Methods:

• Analysis of all-atom molecular dynamics trajectories against crystallographic Protein Data Bank (PDB) structures.
• Development of a PDB-based Statistical Method for reconstructing Cβ atom dynamics.
• Comparison of reconstruction precision with Remo, Pulchra, and Scwrl4 methods.

Main Results:

• Simulated Cβ atom dynamics at physiological conditions can be reconstructed with high precision using backbone Cα and O positions.
• The PDB-based Statistical Method surpasses the precision of Remo and Pulchra.
• Reconstruction precision is comparable to the enhanced Scwrl4 method.

Conclusions:

• The relative positions of Cβ atoms to backbone atoms in dynamic proteins show minimal deviation from static PDB structures.
• Biologically active protein dynamics may adhere to stringent stereochemical constraints similar to folded proteins.
• Results strongly support the development of coarse-grained techniques utilizing reduced coordinate representations.