Automated real-space refinement of protein structures using a realistic backbone move set

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Many biological macromolecule crystals diffract to limited resolution, complicating accurate model building and refinement.
• Current methods for improving low-resolution structural data are often time-consuming and labor-intensive.

Purpose of the Study:

• To present an automated torsional optimization protocol for enhancing moderate- to low-resolution macromolecular structures.
• To improve the quality of structural models derived from crystallography and cryo-electron microscopy (cryo-EM).

Main Methods:

• Developed a protocol combining Protein Data Bank (PDB)-based torsional optimization with real-space refinement.
• Applied the method against electron density maps from crystallography or cryo-EM data.
• Validated the protocol's effectiveness on various protein types, including DNA-binding and membrane proteins.

Main Results:

• Successfully converted moderate- to low-resolution structures into models with improved hydrogen bonding and backbone geometry.
• Achieved better crystallographic R-factors, indicating higher model accuracy.
• Demonstrated applicability to proteins of any size and at different refinement stages.

Conclusions:

• The automated protocol significantly enhances structural model quality from limited-resolution diffraction data.
• This method is broadly applicable to various proteins and can be extended to Nuclear Magnetic Resonance (NMR) and other structural data.
• The protocol aids structural biology studies by improving model quality and reducing refinement time.