Knotify_V2.0: Deciphering RNA Secondary Structures with H-Type Pseudoknots and Hairpin Loops.

Area of Science:

• Computational Biology
• Bioinformatics
• Molecular Biology

Background:

• Accurate prediction of RNA secondary structures is crucial for understanding biological functions.
• The COVID-19 pandemic highlighted the urgent need for advanced RNA structure prediction tools.
• Predicting RNA pseudoknots, especially H-type with bulges and hairpins, remains a significant challenge.

Purpose of the Study:

• To present Knotify_V2.0, a novel framework for predicting RNA secondary structures.
• To specifically enhance the prediction of H-type pseudoknots, including bulges and hairpins.
• To improve upon existing RNA structure prediction methodologies through syntactic pattern recognition.

Main Methods:

• Utilized syntactic pattern recognition techniques.
• Leveraged Context-Free Grammar (CFG) for its expressive capabilities.
• Integrated minimum free energy and maximum base pairing criteria for ambiguity management.

Main Results:

• Knotify_V2.0 demonstrated superior accuracy in predicting core stems compared to state-of-the-art frameworks.
• Successfully identified core base pairing forming ground truth pseudoknots in 70% of sequences.
• Achieved higher true positive counts and lower false negative counts than Knotify, leading to a superior F1-score.

Conclusions:

• Knotify_V2.0 offers enhanced capabilities for predicting complex RNA secondary structures, including pseudoknots.
• The framework shows significant improvements in accuracy, precision, and recall metrics.
• Knotify_V2.0 represents a substantial advancement in RNA structure prediction, with publicly available code.