QRNAS: software tool for refinement of nucleic acid structures.

Area of Science:

• Computational biology
• Structural biology
• Bioinformatics

Background:

• Computational RNA 3D structure models often contain inaccuracies due to simplifications in prediction methods.
• Refinement considering atomic interactions is crucial for improving both local and global model quality.
• Existing methods require enhancement to achieve higher fidelity RNA structural models.

Purpose of the Study:

• To introduce QRNAS, a software tool for fine-grained refinement of nucleic acid structures.
• To extend the AMBER simulation method with additional restraints for enhanced RNA modeling.
• To enable the modeling of diverse nucleic acid structures, including modified residues.

Main Methods:

• QRNAS is presented as an extension of the AMBER simulation method.
• The software incorporates additional restraints for fine-grained refinement.
• QRNAS handles RNA, DNA, chimeras, hybrids, and modified nucleic acid residues.

Main Results:

• QRNAS successfully improved the quality of RNA structural models generated by various methods.
• The software enhanced MolProbity scores for both NMR structures and computational models from RNA-Puzzles.
• Geometry improvements were observed, particularly in base-pair accuracy, though systematic RMSD reduction was not guaranteed.

Conclusions:

• QRNAS effectively refines computational RNA 3D structures, enhancing model quality.
• The tool integrates into computational workflows, improving overall RNA structure prediction accuracy.
• QRNAS demonstrates utility in refining diverse nucleic acid structures, including modified residues.