Semiautomated model building for RNA crystallography using a directed rotameric approach

Area of Science:

• Structural Biology
• Computational Biology
• Biochemistry

Background:

• Crystallographic studies of structured RNA molecules are crucial for understanding cellular processes but face significant challenges.
• Low-resolution electron density maps in RNA crystallography are imprecise and difficult to interpret, hindering accurate structure determination.
• A lack of specialized computational tools for RNA modeling, unlike those for protein crystallography, complicates the model-building process, especially for the RNA backbone.

Purpose of the Study:

• To develop a computational method for accurately building the RNA backbone into electron density maps of intermediate or low resolution.
• To address the difficulties and errors associated with modeling the RNA backbone, which has numerous variable torsion angles.

Main Methods:

• A semiautomated method requiring initial user identification of phosphates and bases in the electron density map.
• Prediction of RNA backbone conformers using RNA pseudotorsions and base-phosphate perpendicular distance.
• Calculation of detailed backbone coordinates that conform to predicted conformers and located structural elements.

Main Results:

• The developed technique accurately builds RNA backbone structures even from imprecise phosphate and base coordinates.
• The method enables accurate backbone modeling without further user intervention after the initial trace.
• A program implementing this methodology is available, with a Coot plugin under development.

Conclusions:

• This novel method significantly improves the accuracy and efficiency of RNA backbone modeling in crystallography.
• The tool addresses a critical gap in computational resources for RNA structural biology.
• The availability of the program and upcoming Coot plugin will aid crystallographers in determining RNA structures.