ScanFold3: a robust and rapid approach for identifying structured RNA targets in genomes and transcriptomes

ScanFold3 speeds structured RNA target identification on sequences up to 128 kb, delivering a 38-fold to 129-fold increase over ScanFold2 while improving PPV on annotated human RNAs.

• Why it matters: High-throughput transcriptomics needs scalable ways to find functional RNA secondary structures, but thermodynamics-based folding often struggles with genomic-scale runtimes.
• What they did: ScanFold3 reformulates consensus structure generation as a minimum perfect weighted matching on a graph, using a Greedy Approximation Algorithm and C++ optimizations for the Fold step.
• The result: Using ScanFold3 outputs as structural constraints for global refolding improves performance versus previous ScanFold versions across all metrics, including the F1 score.

AI summary of the abstract · may contain errors · not medical advice · read the paper.