dsrnascan 0.5.0

A tool for genome-wide prediction of double-stranded RNA structures

dsRNAscan

dsRNAscan is a bioinformatics tool for genome-wide identification of double-stranded RNA (dsRNA) structures. It uses a sliding window approach to detect inverted repeats that can form dsRNA secondary structures, with support for G-U wobble base pairing and ML-based scoring.

Browse human genome results at dsrna.chpc.utah.edu

Install

pip install dsrnascan

Platforms: Linux (Python 3.8+), macOS (Python 3.9+). Windows not supported (use WSL).

Dependencies (auto-installed): biopython, numpy, pandas, ViennaRNA

If ViennaRNA fails via pip, install with conda: conda install -c bioconda viennarna

Usage

# Basic scan (defaults: -w 10000 -s 500 --score 75 -c 4)
dsrnascan input.fasta

# Specific chromosome with 8 CPUs
dsrnascan genome.fasta --only_seq chr21 -c 8

# Scan a specific region
dsrnascan genome.fasta --only_seq chr21 --start 33455482 --end 33655482

# Faster scan with larger step size (less overlap between windows)
dsrnascan genome.fasta -s 5000 -c 16

# Sensitive scan for shorter dsRNAs
dsrnascan sequence.fasta -w 5000 --min_bp 15 --paired_cutoff 60

Parameters

Parameter	Default	Description
-w	10000	Window size (bp)
-s/--step	500	Step size between windows
-c/--cpus	4	Number of CPUs
--min_bp	25	Minimum base pairs required
--score	75	Minimum einverted score
--paired_cutoff	70	Minimum % paired bases
--only_seq	None	Specific chromosome(s) to scan
--start/--end	Full seq	Region coordinates
--forward-only	False	Forward strand only
--reverse-only	False	Reverse strand only
--no-ml	False	Disable ML scoring
--output-dir	Auto	Output directory

Run dsrnascan --help for advanced options (scoring parameters, repeat length limits, folding temperature, etc.).

Output

Results are written to the output directory:

*_merged_results.txt - Tab-delimited predictions with columns:

• Coordinates: Chromosome, Strand, i_start, i_end, j_start, j_end
• einverted: Score, RawMatch, PercMatch, Gaps
• RNAduplex: dG(kcal/mol), base_pairs, percent_paired, longest_helix, eff_i/j_start/end, i_seq, j_seq, structure
• ML scores: stability_model_score, probing_model_score, likely_edited (Yes/No), likely_forms (Yes/No)

*.bp - IGV arc visualization file

*.gff3 - GFF3 with mRNA/exon types so IGV renders the loop between arms as a thin connecting line

*.bedpe - BEDPE format with paired coordinates (one line per dsRNA)

Browse Results

# Interactive viewer with Forna RNA structure visualization
dsrna-browse results_directory/

# With RNA editing site annotations (BED or GFF3)
dsrna-browse results_directory/ --editing-file editing_sites.bed

Citation

If you use dsRNAscan, please cite:

Comprehensive mapping of human dsRNAome reveals conservation, neuronal enrichment, and intermolecular interactions https://doi.org/10.1101/2025.01.24.634786

Additional Tools

overlap_analyzer - Statistical enrichment analysis for genomic features overlapping dsRNA predictions. See overlap_analyzer/README.md. Not included in PyPI package; clone the repo to access.

License

GNU General Public License v3.0 - see LICENSE.

Issues: GitHub Issues

Acknowledgments

• EMBOSS team for the einverted algorithm
• ViennaRNA team for RNA folding algorithms