annotate plasflow.xml @ 1:ea270d0ad0e6 draft

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/plasflow commit 174dafdf949ebf6e6d6f7a314db870daca8a46d6

<tool id="PlasFlow" name="PlasFlow" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="22.01">
    <description>Prediction of plasmid sequences in metagenomic contigs</description>
    <macros>
        <token name="@TOOL_VERSION@">1.1.0</token>
        <token name="@VERSION_SUFFIX@">0</token>
    </macros>
    <requirements>
        <requirement type="package" version="@TOOL_VERSION@">plasflow</requirement>
    </requirements>
    <command detect_errors="exit_code"><![CDATA[
## PlasFlow supports gz only partially 
## https://github.com/smaegol/PlasFlow/blob/45604ff637c65f709dabff63e665fa60896be241/PlasFlow.py#L444
#if $readFile.ext == "fasta.gz"
    gunzip -c '$readFile' > 'reads.fasta' &&
#else
    ln -s '$readFile' 'reads.fasta' &&
#end if

PlasFlow.py
--input 'reads.fasta'
--output 'output'
--threshold '$threshold'
    ]]></command>
    <inputs>
        <param name="readFile" type="data" format="fasta,fasta.gz" optional="true" label="Contig Sequences" />
        <param argument="threshold" type="float" value="0.7" min="0" max="1" label="Threshold for probability filtering"/>
    </inputs>
    <outputs>
        <data name="probability_table" from_work_dir="output" format="tabular" label="${tool.name} on ${on_string}: Probability table" />
        <data name="chromosomes" from_work_dir="output_chromosomes.fasta" format="fasta" label="${tool.name} on ${on_string}: Chromosomes" />
        <data name="plasmids" from_work_dir="output_plasmids.fasta" format="fasta" label="${tool.name} on ${on_string}: Plasmids" />
        <data name="unclassified" from_work_dir="output_unclassified.fasta" format="fasta" label="${tool.name} on ${on_string}: Unclassified" />
    </outputs>
    <tests>
        <test>
            <param name="readFile" value="test.fasta" ftype="fasta"/>
            <param name="threshold" value="0.7"/>
            <output name="probability_table" value="output" compare="re_match"/>
            <output name="chromosomes" value="output_chromosomes.fasta"/>
            <output name="plasmids" value="output_plasmids.fasta"/>
            <output name="unclassified" value="output_unclassified.fasta"/>
        </test>
        <test>
            <param name="readFile" value="test.fasta.gz" ftype="fasta.gz"/>
            <param name="threshold" value="0.7"/>
            <output name="probability_table" value="output" compare="re_match"/>
            <output name="chromosomes" value="output_chromosomes.fasta"/>
            <output name="plasmids" value="output_plasmids.fasta"/>
            <output name="unclassified" value="output_unclassified.fasta"/>
        </test>
    </tests>
    <help><![CDATA[
PlasFlow is a set of scripts used for prediction of plasmid sequences in
metagenomic contigs. It relies on the neural network models trained on full
genome and plasmid sequences and is able to differentiate between plasmids and
chromosomes with accuracy reaching 96%. It outperforms other available solutions
for plasmids recovery from metagenomes and incorporates the thresholding which
allows for exclusion of incertain predictions.
    ]]></help>
    <citations>
        <citation type="doi">10.1093/nar/gkx1321</citation>
    </citations>
</tool>