Physics-based method to validate and repair flaws in protein structures

Area of Science:

• Computational Chemistry
• Structural Biology
• Biophysics

Background:

• Accurate protein structure determination is crucial for understanding biological function.
• Validation of protein conformations and identification of structural flaws remain challenging.
• Nuclear Magnetic Resonance (NMR) spectroscopy provides valuable structural information.

Purpose of the Study:

• To develop and validate a computational method for assessing protein structure quality using chemical shifts.
• To identify backbone or side-chain flaws in protein models.
• To provide optimized torsional angles for refining protein side-chain conformations.

Main Methods:

• Utilized a combination of Carbon-13 alpha ((13)C(α)) and beta ((13)C(β)) chemical shifts computed using density functional theory.
• Employed an ensemble average of chemical shifts over multiple protein conformations for validation.
• Developed the CheShift-2 web server to implement the method and tested side-chain refinement using torsional angles.

Main Results:

• The method achieved approximately 90% sensitivity in validating protein conformations and detecting flaws at the residue level.
• Side-chain refinement using the computed torsional angles led to optimal agreement between observed and computed chemical shifts for ~94% of flaws.
• Refinement did not introduce a significant number of violations in Nuclear Overhauser Effect (NOE)-based distance restraints.

Conclusions:

• The developed method effectively validates protein structures and identifies conformational errors.
• The computed torsional angles facilitate fast and accurate refinement of protein side-chain conformations.
• Integration with other experimental data like NOEs can further enhance structural modeling accuracy.