New restraints and validation approaches for nucleic acid structures in PDB-REDO

Area of Science:

• Structural Biology
• Biochemistry
• Computational Biology

Background:

• Macromolecular structure model quality relies on accurate refinement and validation.
• Existing software predominantly focuses on protein structures, lacking features for nucleic acids.
• Nucleic acid structures require specialized approaches for accurate modeling.

Purpose of the Study:

• To introduce novel nucleic acid-specific refinement and validation methods within the PDB-REDO software.
• To enhance the quality of structure models containing DNA and RNA.
• To provide improved tools for the structural biology community.

Main Methods:

• Implementation of a new restraint model in PDB-REDO utilizing noncovalent geometries (hydrogen bonding, base-pair stacking) as refinement targets.
• Development of new validation routines, including a metric for Watson-Crick base-pair geometry normality (ZbpG).
• Application of the enhanced PDB-REDO pipeline to the entire Protein Data Bank (PDB).

Main Results:

• The PDB-REDO pipeline with the new restraint model positively impacted the quality of nucleic acid-containing structure models.
• Demonstrated improvements in the geometry of specific nucleic acid structures within the PDB.
• The new methods provide more accurate representations of nucleic acid conformations.

Conclusions:

• The developed nucleic acid-specific approaches in PDB-REDO significantly improve macromolecular structure model quality.
• These advancements offer valuable tools for researchers working with nucleic acid structures.
• The enhanced PDB-REDO models and pipeline are publicly available for broader scientific use.