<tool id="virannot_otu" name="virAnnot OTU" version="1.0.1+galaxy0" profile="21.05">
    <description>create viral OTUs based on RPS and Blast annotations</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="xrefs"/>
    <expand macro="requirements"/>
    <required_files>
        <include path="otu.py" />
        <include path="seek_otu.R" />
        <include path="rps2tree_html.py" />
    </required_files>
    <command detect_errors="aggressive"><![CDATA[

@HEADLESS@
    #if $blast_files
        #for $blast_file in $blast_files
            ln -fs '$blast_file' '$blast_file.element_identifier' &&
        #end for
    #end if 
    #for $rps_file in $rps_files
        ln -fs '$rps_file' '$rps_file.element_identifier' &&
    #end for
    #for $fasta_file in $fasta_files
        ln -fs '$fasta_file' '$fasta_file.element_identifier' &&
    #end for

    python '$__tool_directory__/otu.py'
    #if $blast_files
        #for $blast_file in $blast_files
            -b '$blast_file.element_identifier'
        #end for
    #end if 
    #for $rps_file in $rps_files
        -r '$rps_file.element_identifier'
    #end for
    #for $fasta_file in $fasta_files
        -f '$fasta_file.element_identifier'
    #end for
    -p '$percentage'
    -vp '$viral_portion'
    -mpl '$min_prot_length'
    -tp '$__tool_directory__/'
    -o otu_output

    && zip -r out.zip otu_output

    ]]></command>
    <inputs>
        <param type="data" name="blast_files" format="tabular" label="Blast results file with taxonomy" multiple="true" optional="true"
            help="Give one file per sample. If a file is missing, give none." />
        <param type="data" name="rps_files" format="tabular" label="RPS results file" multiple="true" />
        <param type="data" name="fasta_files" format="fasta" label="Contigs file" multiple="true" />
        <param type="integer" name="percentage" label="Similarity threshold percentage for OTUs cutoff" value="90" min="10" max="100" />
        <param type="select" name="viral_portion" label="Minimun portion of viral sequences in RPS domain to be included">
            <option value="0.1">0.1</option>
            <option value="0.2">0.2</option>
            <option value="0.3" selected="true">0.3</option>
            <option value="0.4">0.4</option>
            <option value="0.5">0.5</option>
        </param>
        <param type="select" name="min_prot_length" label="Minimum query protein length">
            <option value="50">50</option>
            <option value="80">80</option>
            <option value="100" selected="true">100</option>
            <option value="150">150</option>
            <option value="200">200</option>
        </param>
    </inputs>
    <outputs>
        <data format="xlsx" name="otu_table" label="${tool.name} on ${on_string}: otu_stats" from_work_dir="otu_output/otu_stats.xlsx" />
        <data format="zip" name="archive_output" label="${tool.name} on ${on_string}: zip of the complete output" from_work_dir="out.zip" />
    </outputs>
    <tests>
        <test>
            <param name="blast_files" value="otu_s1_tblastx.tab,otu_s2_tblastx.tab"/>
            <param name="rps_files" value="otu_s1_rps.tab,otu_s2_rps.tab"/>
            <param name="fasta_files" value="otu_s1.fa,otu_s2.fa"/>
            <param name="percentage" value="90"/>
            <param name="viral_portion" value="0.3"/>
            <param name="min_prot_length" value="100"/>
            <!-- check created zip -->
            <output name="archive_output">
                <assert_contents>
                    <has_archive_member path=".*\/index.html" />
                    <has_archive_member path=".*\/pfam00680_RdRP_1/seq_aligned.final_tree.fa" />
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[

This module takes as input rps file from rps2tsv, blast file from blastxml_to_tabular [optional] and fasta contig file [optional].

If no blast file is given, OTUs will not have an associated annotation i.e the OTUs will be set as unknown.
Files are processed in alphabetical order. Make sure that the files (blast, rps and contig) for a sample are named in the same way,
otherwise they will be disassociated.

The standard similarity threshold is 90% [default value] which is a proxy to viral species compared with the ICTV classification.

    ]]></help>
    <expand macro="citations" />
</tool>