<tool id="distance" name="Quasispecies Distance" version="0.4.2">
    <description>Calculate the evolutionary distance between viral quasispecies.</description>
    <requirements>
          <requirement type="package" version="0.4.2">quasitools</requirement>
    </requirements>
    <command detect_errors="exit_code"><![CDATA[

        #for $bam_file in $bam_files
        ln -s $bam_file '${bam_file.name}' &&
        ln -s $bam_file.metadata.bam_index '${bam_file.name}.bai' &&
        #end for

        quasitools distance $ref_file

        #for $bam_file in $bam_files
        '${bam_file.name}'
        #end for

        -o matrix.csv

        $normalize

        #if $low_coverage.coverage_selector == 'keep':
            -k
        #else if $low_coverage.coverage_selector == 'remove':
            -r
        #else if $low_coverage.coverage_selector == 'truncate':
            -t
        #end if

        #if $select_range.manual_input == 'select_range':
            -s $select_range.start
            -e $select_range.end
        #end if

        $output_similarity

    ]]></command>
    <inputs>
        <param name="ref_file" type="data" format="fasta" optional="false" label="Reference file" />
        <param name="bam_files" type="data" format="bam" optional="false" multiple="true" label="BAM files" min="2" help="At least two BAM files are required to perform quasispecies distance comparison."/>
        <param name="normalize" type="boolean" truevalue="-n" falsevalue="-dn" checked="True" label="Normalize pileups" help="Normalize read count data so that the read counts per 4-tuple (A, C, T, G) sum to one."/>
        <param name="output_similarity" type="boolean" truevalue="-os" falsevalue="-od" checked="False" label="Output cosine similarity instead of angular cosine distance." help="Cosine similarity is not a distance metric." />
        <conditional name="low_coverage">
            <param name="coverage_selector" type="select" label="Choose how to treat zero coverage positions.">
                <option value="remove" selected = "true">Remove all zero coverage positions.</option>
                <option value="keep">Keep all zero coverage positions.</option>
                <option value="truncate">Truncate zero coverage positions at start and end of pileup.</option>
            </param>
            <when value="remove">
            </when>
            <when value="keep">
            </when>
            <when value="truncate">
            </when>
        </conditional>
        <conditional name="select_range">
            <param name="manual_input" type="select" label="Select pileup range." help="Currently only supports selection of custom start and end position for one single reference.">
                <option value="keep_range" selected = "true">Use original pileup for distance calculation.</option>
                <option value="select_range">Manually select pileup's start and end position for distance calculation.</option>
            </param>
            <when value="select_range">
                <param name="start" type="integer" optional="false" min="1" value="1" label="Select one-indexed start base position." help="Positions before start position will be ignored in the distance calculation." />
                <param name="end" type="integer" optional="false" min="1" value="1" label="Select one-indexed end base position." help="Positions after end position will be ignored in the distance calculation." />
            </when>
            <when value="keep_range">
            </when>
        </conditional>
    </inputs>
    <outputs>
        <data format="csv" name="matrix" from_work_dir="matrix.csv" />
    </outputs>
    <tests>
        <test>
            <param name="ref_file" value="hxb2_pol.fas" />
            <param name="bam_files" value="quasi1.bam,quasi2.bam" />
            <param name="normalize" value="-n" />
            <param name="coverage_selector" value="truncate" />
            <param name="manual_input" value="keep_range" />
            <param name="output_similarity" value="-od" />
            <output name="matrix" ftype="csv" >
                <assert_contents>
                    <has_text_matching expression="Quasispecies"/>
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[

Quasispecies Distance
=====================

This tool determines the cosine relatedness between viral quasispecies, reporting either angular cosine distance or cosine similarity as measures of relatedness. These measures of relatedness should be understood as approximations for evolutionary distance and might be useful for performing clustering.

------
Inputs
------

1. A reference FASTA.
2. At least two BAM files generated in coordination with the above reference.

----------
Parameters
----------

1. Normalize pileup (True or False).
2. Select custom start and end position. Currently only supports selection of custom start and end position for one single reference.
3. Choose how to treat zero coverage regions (keep, remove all, or truncate).

-------
Outputs
-------

A CSV file containing a matrix of similarities or distances, as requested by the user.

    ]]></help>
    <citations>
        <citation type="bibtex">
            @misc{GitHubquasitoolsdistance,
            title = {quasitools distance},
            publisher = {phac-nml},
            journal = {GitHub repository},
            url = {https://github.com/phac-nml/quasitools},
            }
        </citation>
    </citations>
</tool>