Studying the Protein Thermostabilities and Folding Rates by the Interaction Energy Network in Solvent

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Protein characteristics like folding rate, thermostability, and allostery are governed by residue interaction networks.
• Interactions can be quantified by distance (simple but less rigorous) or energy (more precise, especially with solvent effects).

Purpose of the Study:

• To apply and validate an energy decomposition method for constructing protein interaction energy (IE) networks.
• To demonstrate the utility of IE networks in predicting key protein properties and mechanisms.

Main Methods:

• Utilized an existing energy decomposition method based on the Poisson-Boltzmann equation solver.
• Accelerated calculations using Graphics Processing Units (GPUs) for enhanced performance.
• Constructed and analyzed IE networks for four distinct applications.

Main Results:

• Energy-based contact order showed a stronger correlation (PCC=0.839) with protein folding rates than distance-based contact order (PCC=0.784).
• Thermophilic proteins generally exhibited lower interaction energies (IEs) in solvent compared to mesophilic proteins, indicating IE as a thermostability indicator.
• IE network analysis successfully predicted key residues for insulin dimer formation, aligning with experimental findings.
• A novel method (APFN) based on IE networks accurately predicted the allosteric pathway for CheY protein, consistent with NMR spectroscopy results.

Conclusions:

• The IE network in solvent is a reliable tool for characterizing proteins.
• This energy-based approach provides deeper insights into protein folding, stability, and allosteric regulation.
• GPU acceleration significantly enhances the computational efficiency of IE network construction.