The PDB_REDO server for macromolecular structure model optimization.

Area of Science:

• Structural Biology
• Biochemistry
• Crystallography

Background:

• Model refinement and validation are crucial before submitting crystallographic data to the Protein Data Bank (PDB).
• Ensuring high-quality PDB archives relies on rigorous validation against geometrical criteria and diffraction data.
• The PDB_REDO procedure focuses on constructive validation, optimizing refinement and proactively rebuilding models for accurate electron density interpretation.

Purpose of the Study:

• To introduce a web server for the PDB_REDO procedure, enabling automated, consistent optimization of crystallographic structure model refinement.
• To assist practicing crystallographers in enhancing their models before PDB submission.
• To implement advanced validation criteria and innovative result presentation for improved model inspection.

Main Methods:

• Implementation of a web server for automated refinement using REFMAC and partial model rebuilding.
• Integration of additional steps in the PDB_REDO pipeline, including resolution limit testing and k-fold cross-validation.
• Inclusion of new validation criteria such as density-fit metrics (EDSTATS) and ligand validation (YASARA).

Main Results:

• The PDB_REDO web server provides thorough, consistent, and automated optimization of the refinement procedure.
• Additional validation steps and metrics enhance the quality assessment of crystallographic models.
• Innovative presentation of results and auto-generated Coot scripts facilitate user-guided model improvement.

Conclusions:

• The PDB_REDO web server significantly improves crystallographic structure models prior to PDB submission.
• Automated refinement and advanced validation contribute to higher quality structural data in the PDB.
• The tool empowers crystallographers to refine and validate their models effectively, ensuring data integrity.