A tool to search ShortStack sRNA annotations against miRBase

This tool is in development, it is likely that many of the options and outputs may change. There are many bugs still being worked out.

ShortStack is a powerful tool for sRNA-seq analysis. It produces an annotation, including high-precision predictions of miRNAs. However, precision comes with a trade-off of sensitivity and loci which contain miRNAs may not always be identified as such. It can also be useful to find family names associated with a sRNA locus. This tool is meant to bridge that gap and help identify loci that might be overlooked otherwise.

This tool uses 3 approaches to search for miRBase miRNA loci all with varying degrees of sensitivity.

• Exact mature miRNA matches between a submitted miRNAs and MajorRNAs from ShortStack (the most precise).
• Genomic overlap of BLAT alignment of submitted hairpins vs ShortStack annotated loci.
• BLAST alignment of ShortStack annotated loci vs submitted hairpins.

Using these 3 approaches, it can find what family members are most similar to a locus and give a strong inference what miRNA families might be contained in a locus.

• MiRBase is not perfect. As of now this tool includes low-confidence miRNAs. These annotations should be considered miRNAs at the user's peril - many of them are likely false.
• This will likely be only specific to the family of a miRNA. Any species information or paralog information should be disregarded - they mean nothing in the context of this tool. So, a hit that said Ath-miR166b when you are working in A. lyrata, is not evidence of horizontal gene transfer... You should only trust this as far as this is a miR166-family miRNA.

All of these software must be executable and in the $PATH variable:

• samtools
• bedtools
• blastn
• blat
• (python3)

Past these, there is no installation. Simply run the stack-to-base.py script. To run this executable from anywhere, add the repo to your $PATH variable in your bash_rc or bash_profile, and confirm the tool is executable:

chmod +x ./stack-to.base.py

This tool uses hairpin and mature sequences from mirbase (version?). Both of these files have been supplied in the github repo and stack-to-base.py should recognize their location if the script is in the repo directory. Otherwise, you can specify their location manually.

Check the help document for syntax information:

stack-to-base.py -h

The tool simply takes the Results.txt output from a ShortStack run and the genome used to perform the annotation. An output name is supplied to handle intermediate files.

stack-to-base.py -r example/Results.txt -g ~/+Genomes/Plant_genomes/Solyc.chromosomes.4.00.fa -o test_output

A possibly better way to use this tool is to include non-mirbase annotations as well, for example those found in Lunardon et al 2020 hosted on the plant small RNA genes server. These can be submitted as more arguments for the -m (Mature) and -p (hairPin) options.

stack-to-base.py -r example/Results.txt -g ~/+Genomes/Plant_genomes/Solyc.chromosomes.4.00.fa -o test_output \
    -m miRBase21.mature.fa Lunardon2020.mature.fa \
    -p miRBase21.hairpin.fa Lunardon2020.hairpin.fa