Protein folding by restrained energy minimization and molecular dynamics

Area of Science:

• Computational Biology
• Protein Structure Prediction
• Biophysics

Background:

• Determining accurate protein conformations is crucial for understanding biological function.
• Existing methods for protein structure prediction face challenges in achieving native-like folded states.

Purpose of the Study:

• To develop and validate a computational method for calculating native-like folded conformations of proteins.
• To explore the relationship between energy, conformation, and deviation from experimental structures.

Main Methods:

• Utilized soft-atom restrained energy minimization on 25 random starting structures.
• Incorporated restraints for disulfide bridges and main-chain hydrogen bonds defining secondary structure.
• Employed molecular dynamics followed by energy minimization for conformational annealing.
• Analyzed conformations using writhing numbers, torsion angle distributions, and accessible surface areas.

Main Results:

• Calculated diverse protein conformations with varying energies and root-mean-square deviations (RMSD).
• Established a correlation between lower energy and closer proximity to the experimental X-ray structure.
• Achieved a best conformation with an RMSD of 3 Å, matching the native threading.

Conclusions:

• The developed computational approach effectively calculates low-energy, native-like protein conformations.
• The method's success in predicting the bovine pancreatic trypsin inhibitor structure highlights its potential.
• This technique offers broad applicability for modeling protein conformations under various restraints.