Quantum-mechanics-derived 13Calpha chemical shift server (CheShift) for protein structure validation.

Area of Science:

• Computational Chemistry
• Structural Biology
• Bioinformatics

Background:

• Accurate prediction of protein structure is crucial for understanding biological function.
• Nuclear Magnetic Resonance (NMR) spectroscopy and X-ray crystallography provide experimental structural data.
• Carbon-13 alpha ((13)C(alpha)) chemical shifts are sensitive probes of protein backbone conformation.

Purpose of the Study:

• To develop a computational server, CheShift, for predicting (13)C(alpha) chemical shifts in protein structures.
• To evaluate the accuracy and sensitivity of CheShift predictions compared to existing methods.
• To establish CheShift as a tool for protein structure validation using experimental chemical shift data.

Main Methods:

• Generation of over 696,000 protein conformations based on torsional angles (phi, psi, omega, chi1, chi2).
• Computation of (13)C(alpha) chemical shifts using Density Functional Theory (DFT) with basis set extrapolation.
• Validation using X-ray and NMR-derived protein structures, including decoy sets, and comparison with SHIFTS, SHIFTX, SPARTA, and PROSHIFT.

Main Results:

• CheShift accurately predicts (13)C(alpha) chemical shifts, showing high correlation coefficients (R) and low root-mean-square deviations (rmsd).
• Comparative analysis reveals CheShift as the most sensitive server for detecting subtle structural differences.
• CheShift effectively distinguishes between various protein models and aids in structure validation.

Conclusions:

• CheShift provides a sensitive and accurate method for predicting protein (13)C(alpha) chemical shifts.
• The server is a valuable tool for validating protein structures determined by X-ray crystallography and NMR.
• CheShift is accessible as a web server for the scientific community.