An Aggregation Method to Identify the RNA Meta-Stable Secondary Structure and its Functionally Interpretable...

Area of Science:

• Molecular Biology
• Computational Biology
• Bioinformatics

Background:

• RNA molecules perform essential cellular functions via their complex structures.
• Experimental methods for RNA structure determination are low-throughput.
• Computational tools often overlook RNA structure ensembles and functional interpretation.

Purpose of the Study:

• To develop a novel computational method for identifying functionally interpretable RNA structure ensembles.
• To predict the meta-stable structure, representing the most frequent functional RNA conformation.
• To enable the deciphering of functional hypotheses from predicted RNA structures.

Main Methods:

• Development of a novel computational approach to predict RNA structure ensembles.
• Identification of meta-stable RNA structures from predicted ensembles.
• Validation using yeast and human test sets, including functional inference and biological case studies.

Main Results:

• The proposed method accurately predicts meta-stable RNA structures, outperforming existing tools on yeast data.
• Inferred functional aspects achieved a micro-averaging F1 score of 0.92 upon manual verification.
• The method demonstrated applicability across species (yeast, human) and robustness against sequence length variations.

Conclusions:

• The novel method effectively predicts functionally relevant RNA structure ensembles and meta-stable conformations.
• It provides testable hypotheses for RNA function, as illustrated by the ASH1-E1 element example.
• The approach is broadly applicable and overcomes limitations of current RNA structure prediction tools.