TFold

TFold predicts non-coding RNA secondary structures with a focus on identifying microRNA (miRNA) precursors (pre-miRNAs) within genomic sequences.

Key Features:

• Algorithmic Efficiency: Employs an approach that searches for approximate miRNA hairpins to reduce the candidate search space and decrease computational complexity.
• Speed and Performance: Optimized for rapid processing, reported to process a 1 MB genomic sequence in approximately 30 seconds, compared with CID-miRNA (55 hours), miRPara (20 hours), and VMir (30 minutes).
• Sensitivity and Selectivity: Demonstrates high sensitivity and selectivity in identifying pre-miRNAs, outperforming several established methods across various genomic sequences.
• Ab-initio Methodology: Predicts pre-miRNA structures directly from sequence data without relying on prior RNA-structure knowledge.

Scientific Applications:

• Genomic research: Rapid prediction of non-coding RNA secondary structures within genomic sequences.
• Next-generation sequencing analysis: Efficient processing of large genomic datasets generated by next-generation sequencing technologies.
• Novel miRNA discovery and validation: Facilitates identification and validation of novel pre-miRNAs from genomic data.
• Molecular biology, genetics, and bioinformatics studies: Supports investigations into miRNA-mediated gene regulation and non-coding RNA biology.

Methodology:

The methodology comprises two explicitly stated computational steps: approximation of miRNA hairpins to narrow candidate regions, followed by reconstitution of full pre-miRNA structures from those candidates.