The Multiscale Ernwin/SPQR RNA Structure Prediction Pipeline.

Area of Science:

• Biochemistry
• Computational Biology
• Structural Biology

Background:

• RNA molecules perform diverse biological functions beyond protein synthesis, dictated by their 3D structures.
• Predicting RNA structure from sequence is crucial but challenging, especially for long noncoding RNAs (lncRNAs).
• Existing computational tools struggle with the complexity and size of lncRNAs (>200 nucleotides).

Purpose of the Study:

• To develop and describe a novel multiscale computational approach for predicting the 3D structures of lncRNAs.
• To integrate experimental data, specifically SAXS (Small-Angle X-ray Scattering), into the structure prediction workflow.
• To refine and validate the proposed methodology using a specific lncRNA, Braveheart.

Main Methods:

• A hierarchical, multiscale modeling strategy was employed.
• The approach couples two coarse-grained models: Ernwin (helix-based, global arrangement) and SPQR (nucleotide-centered, refinement).
• SAXS and secondary structure experimental data were incorporated to guide and refine the structure prediction.

Main Results:

• The methodology was successfully applied to predict the structure of the Braveheart lncRNA.
• The multiscale approach, integrating experimental data, yielded a refined all-atom structure.
• The study demonstrates the feasibility of accurately modeling complex lncRNA structures.

Conclusions:

• The described multiscale approach effectively predicts lncRNA structures by combining computational modeling with experimental data.
• This method addresses the challenges posed by the size and complexity of lncRNAs.
• The refined all-atom structure of Braveheart lncRNA serves as a validation of the proposed technique.