Decomposition of Proteins into Dynamic Units from Atomic Cross-Correlation Functions

Area of Science:

• Computational biology
• Biophysics
• Protein dynamics

Background:

• Proteins are complex molecules whose functions depend on their dynamic behavior.
• Traditional methods often decompose proteins based on static structural properties.
• Understanding protein dynamics is crucial for deciphering biological mechanisms.

Purpose of the Study:

• To introduce a new method for clustering protein atoms based on dynamic correlations.
• To provide a coarse-grained protein description using dynamically independent subunits.
• To complement existing structure-based domain decomposition techniques.

Main Methods:

• Utilizing molecular dynamics (MD) simulations to compute interatomic distance correlation functions.
• Analyzing the correlation times of these functions to identify dynamically correlated domains.
• Clustering protein residues based on effective correlation times, independent of structural features.

Main Results:

• Demonstrated a method for decomposing proteins into dynamically correlated domains.
• Showcased the application of this method on a prototypal protein structure.
• Highlighted the distinct nature of this dynamic-based clustering compared to structure-based methods.

Conclusions:

• The proposed clustering method offers a novel, dynamics-centric approach to protein analysis.
• This method provides a complementary view to structure-based domain decomposition.
• It enables a coarse-grained representation of proteins based on functional dynamics.