SnapT - Small NcRNA Annotation Pipeline for (meta)Transcriptomic data

SnapT is a Small non-coding RNA annotation pipeline for Transcriptomic or metatranscriptomic data. SnapT leverages transcriptomic or metatranscriptimic data to find, annotate, and quantify intergenic and anti-sense sRNA transcripts. To do this, SnapT aligns reads from a stranded RNAseq experiment to the reference (meta)genome, and then assembles the reads into transcripts. The transcripts are then intersected with the genome annotation as well as open reading frames to select for only transcripts that fall on non-coding regions, and further filtered to produce a final set of predicted small ncRNAs:

Intergenic transcripts must be at least 30nt away from any gene or ORF on both strands
Antisense transcripts must be 30nt away from any gene on their strand, but overlap with a gene on the opposite strand by at least 10nt.
Small peptides (<100nt) are not counted as a genes if they are encoded in a transcript that is more than 3 times their length.
The non-coding transcripts cannot contain any ORF greater than 1/3 of their length.
Predicted non-coding transcripts near contig edges are discarded due to mis-annotation potential.
Small ncRNAs must be between 50nt and 500nt in length
The transcripts must not have signifficant homology with any protein in the NCBI_NR database (query cover>30%, Bitscore>50, evalue<0.0001, and identity>30%).
The transcripts must not have homology with any model (except for small RNAs) in the Rfam non-coding RNA database.

SNAPT PIPELINE WORKFLOW

SnapT small ncRNA annotation pipeline

INSTALLATION

Conda installation:

To start, download miniconda2 and install it:

 wget https://repo.continuum.io/miniconda/Miniconda2-latest-Linux-x86_64.sh #FOR LIXUX
 bash Miniconda2-latest-Linux-x86_64.sh

Then simply install SnapT from the ursky conda channel (supports Linux64 and OsX):

 conda install -c ursky snapt

Download and index the NCBI NR protain database (you will need to input the nr.dmnd index into snapt with the -D or --nr option). The downloaded file should be 43GB (108GB unzipped), and indexing should take under an hour.

wget ftp://ftp.ncbi.nlm.nih.gov/blast/db/FASTA/nr.gz
gunzip nr.gz
mv nr nr.faa
diamond makedb --in nr.faa -d nr

Finally, download and index the Rfam non-coding RNA database (you will need to input the Rfam.cm database into snapt with the -R or --rfam option). The downloaded file should be 35MG (245MB unzipped), and indexing should take a few seconds.

wget ftp://ftp.ebi.ac.uk/pub/databases/Rfam/14.1/Rfam.cm.gz
gunzip Rfam.cm.gz
cmpress Rfam.cm

Manual installation:

You may want to manually install SnapT if you want better control over your environment, if you are installing on non-conventional system, or you just really dislike conda. In any case, you will need to manually install the relevant prerequisite programs. When you are ready, download or clone this ripository and add the SnapT/bin/ directory to to the $PATH or copy the SnapT/bin/ contents into a directory that is under PATH. Thats it!

USAGE

Example run of Snapt:

 snapt -1 READS/ALL_1.fastq -2 READS/ALL_2.fastq -g metagenomic_assembly.fasta -a metagenomic_assembly.gff -l 3000 -o SNAPT_OUT -t 48 --nr ../DATABASES/NCBI_nr/nr.dmnd --rfam ../DATABASES/rfam/Rfam.cm

Help message

Usage: SnapT [options] -1 reads_1.fastq -2 reads_2.fastq -g genome.fa -o output_dir

SnapT options:
    -1 STR      forward transcriptome (or metatranscriptome) reads
    -2 SRT      reverse transcriptome (or metatranscriptome) reads (optional)
    -g STR      genome (or metagenome) fasta file
    -a STR      genome (or metagenome) annotation gtf/gff file (optional, but recommended)
    -l INT      minimum contig length (default=1000) for ncRNA annotation
    -o STR          output directory
    -t INT      number of threads (default=1)
    -d STR      NCBI_nr protein database DIAMOND index (see installation instructions for details)
    -rfam       run analysis of ncRNAs detected against rfam database (rfam.cm) to annotate known ncRNAs/sRNAs (see installation instructions for details)

Aligment options:
    -r STR      rna-strandness: R or F for single-end, RF or FR for paired-end (default=FR)
    -I INT      min insert size (default=0)
    -X INT      max insert size (default=500)
    -m INT      gap distance to close transcripts (default=50)

    --version | -v  show current SnapT version

Citing SnapT

SnapT and its validation is outlined in the publication "Regulatory Noncoding Small RNAs Are Diverse and Abundant in an Extremophilic Microbial Community", published in mSystems.

Acknowledgements

Authors of pipeline: Gherman Uritskiy and Diego Gelsinger

Principal Investigator: Jocelyne DiRuggiero

Institution: Johns Hopkins, Department of Cell, Molecular, Developmental Biology, and Biophysics

All feedback is welcome! For errors and bugs, please open a new Issue thread on this github page, and we will try to get things patched as quickly as possible. Please include the version of SnapT you are using (run snapt -v). For general questions about the conda impementation of this software, contact Gherman Uritskiy at guritsk1@jhu.edu. For general questions or suggestions about the pipeline itself, contact Diego Gelsinger at drg2165@cumc.columbia.edu.

ursky / SnapT

readme