Spectral analysis of molecular dynamics simulations on PDZ: MD sectors

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Protein sectors are defined as sparse subsets of amino acid residues forming cooperative networks essential for protein stability, ligand binding, and allosterism.
• Statistical Coupling Analysis (SCA) has been the primary method for identifying protein sectors using evolutionary conservation data from multiple sequence alignments.
• SCA-derived sectors have demonstrated correlations with functional properties and allosteric mechanisms in various protein systems.

Purpose of the Study:

• To explore the application of physics-based molecular dynamics (MD) simulations for identifying protein sectors.
• To validate the utility of MD-derived sectors by comparing them with established methods and experimental data.

Main Methods:

• All-atom MD simulations were conducted for the PSD95-PDZ3 protein for 500 nanoseconds using AMBER.
• Covariance matrices were generated from MD-derived residue displacement and non-bonded interaction energy fluctuations.
• Spectral analysis was applied to these MD-calculated matrices, mirroring the approach used in SCA.

Main Results:

• Sectors identified through MD simulations showed concordance with those identified by SCA.
• The MD sector composition aligned with results from site-specific mutagenesis and known allosteric behavior.
• This indicates that MD simulations can effectively capture the cooperative networks characteristic of protein sectors.

Conclusions:

• Molecular dynamics (MD) sectors provide a viable and complementary protocol for analyzing protein dynamics and identifying functional networks.
• This approach offers insights into the physical underpinnings of protein allosterism and stability.
• MD-based sector analysis can be a powerful tool for understanding protein function from simulation data.