Accurate SHAPE-directed RNA secondary structure modeling, including pseudoknots.

Area of Science:

• Computational Biology
• RNA Structure Prediction
• Biochemistry

Background:

• RNA pseudoknots are crucial functional motifs found in catalytic RNAs, riboswitches, and viral elements.
• Traditional RNA structure prediction algorithms often exclude pseudoknots due to modeling complexity and computational challenges.
• Accurate prediction of RNA secondary structures, including pseudoknots, is essential for understanding RNA function.

Purpose of the Study:

• To develop a robust and concise secondary structure modeling approach for RNA, incorporating pseudoknots.
• To integrate experimental chemical probing data with thermodynamic energy models for improved prediction accuracy.
• To accurately predict both secondary structures and pseudoknots in a diverse set of RNA molecules.

Main Methods:

• Combined selective 2'-hydroxyl acylation analyzed by primer extension (SHAPE) experimental data with a thermodynamic energy model.
• Developed a dynamic programming algorithm for iterative refinement of RNA structures.
• Modeled the entropic cost associated with single pseudoknot formation.

Main Results:

• The developed method accurately predicted secondary structures and pseudoknots for 21 challenging RNAs (34-530 nt).
• Achieved an average prediction accuracy of 93% for known base pairs.
• Successfully identified all pseudoknots in well-folded RNA molecules within the tested set.

Conclusions:

• The integrated approach of experimental data and thermodynamic modeling provides accurate RNA secondary structure prediction, including pseudoknots.
• This method overcomes limitations of existing algorithms by effectively modeling complex RNA folds.
• The findings enable more reliable prediction of functionally critical RNA structures.