<tool id="ivar_removereads" name="ivar removereads" version="@TOOL_VERSION@+galaxy0" profile="@PROFILE@">
    <description>Remove reads from trimmed BAM file</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements"/>
    <expand macro="version_command"/>
    <command detect_errors="exit_code"><![CDATA[
        cp '$input_bed' binding_sites.bed &&
        python '$__tool_directory__/sanitize_bed.py' binding_sites.bed &&
        #if $amplicons.computed == "yes"
            python '$__tool_directory__/write_amplicon_info_file.py' binding_sites.bed amplicon_info.tsv &&
        #else
            ln -s '$amplicon_info' amplicon_info.tsv &&
        #end if
        ivar getmasked
        -i '$variants_tsv' -b binding_sites.bed -f amplicon_info.tsv -p masked_primers &&

        python '$__tool_directory__/completemask.py' masked_primers.txt amplicon_info.tsv &&
        ln -s '$input_bam' sorted.bam &&
        ln -s '${input_bam.metadata.bam_index}' sorted.bam.bai &&

        ivar removereads
        -i sorted.bam
        -b binding_sites.bed
        -p removed_reads.bam
        -t masked_primers.txt
    ]]></command>
    <inputs>
        <param name="input_bam" argument="-i" type="data" format="bam"
        label="Bam input"
        help="BAM dataset, preprocessed with ivar trim, to remove reads from"/>
        <param name="variants_tsv" type="data" format="tabular"
        label="Variants input"
        help="This dataset will be scanned for variants that affect primer binding sites and needs to be in tabular format with affected chromosome names in the first, and positions in the second column. If there is a header line, the name of the second column should be POS."/>
        <param name="input_bed" argument="-b" type="data" format="bed" label="Primer binding sites information"
        help="The same six-column BED dataset that served as input to ivar trim"/>
        <conditional name="amplicons">
            <param name="computed" type="select" label="Compute amplicon info from BED file" help="For suitable primer binding site datasets amplicon info can be computed directly (see tool help below). For others you will need to provide an extra amplicon info dataset.">
                <option value="yes" selected="true">Yes</option>
                <option value="no">No</option>
            </param>
            <when value="yes" />
            <when value="no">
                <param name="amplicon_info" type="data" format="tabular"
                label="Primer to amplicon assignment info"
                help="This input should consist of one line per amplicon with the tab-separated names of all primers used to generate that amplicon."/>
            </when>
        </conditional>
    </inputs>
    <outputs>
        <data name="output_bam" format="bam" label="${tool.name} on ${on_string}" from_work_dir="removed_reads.bam"/>
    </outputs>
    <tests>
        <test>
            <param name="input_bam" value="zika/Z52_a.trimmed.sorted.bam"/>
            <param name="variants_tsv" value="zika/primers_Z52_consensus.tsv"/>
            <param name="input_bed" value="zika/db/zika_primers_consensus.bed"/>
            <conditional name="amplicons">
                <param name="computed" value="no"/>
                <param name="amplicon_info" value="zika/db/pair_information.tsv"/>
            </conditional>
            <assert_stdout>
                <!-- check that primer pairs are picked up correctly
                from the amplicon info file -->
                <has_text text="400_18_out_L**&#009;400_18_out_R**&#009;400_23_out_L&#009;400_23_out_R&#009;400_32_out_L&#009;400_32_out_R"/>
            </assert_stdout>
            <output name="output_bam" ftype="bam">
                <assert_contents>
                    <has_size value="3130923" delta="1000"/>
                </assert_contents>
            </output>
        </test>
        <test>
            <param name="input_bam" value="zika/Z52_a.trimmed.sorted.bam"/>
            <param name="variants_tsv" value="zika/primers_Z52_consensus.tsv"/>
            <param name="input_bed" value="zika/db/zika_primers_consensus.bed"/>
            <conditional name="amplicons">
                <param name="computed" value="yes"/>
            </conditional>
            <assert_stdout>
                <!-- check that primer pairs are picked up correctly
                from the autogenerated amplicon info file -->
                <has_text text="400_18_out_L**&#009;400_18_out_R**&#009;400_23_out_L&#009;400_23_out_R&#009;400_32_out_L&#009;400_32_out_R"/>
            </assert_stdout>
            <output name="output_bam" ftype="bam">
                <assert_contents>
                    <has_size value="3130923" delta="1000"/>
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[
This Galaxy tool combines the functionality of ``ivar getmasked`` and
``ivar removereads``. No separate ``ivar getmasked`` step is required when
using this tool.

The wrapper takes as input a BAM dataset of aligned and sorted reads, from
which the primers listed in the primer binding sites BED input have been
trimmed with ``ivar trim``.

From this input it will remove reads that come from amplicons that have been
generated with one or more primers that may have been affected in their binding
by variants listed in the variants input file.
To do its job, the needs to know which primers work together to form an
amplicon. The tool can try to deduce this info from the names of the primers
found in the primer info dataset. This will require a primer naming scheme
following the regex pattern::

  .*_(?P<amplicon_number>\d+).*_(?P<primer_orientation>L(?:EFT)?|R(?:IGHT)?)

*i.e.*, the following schemes will work (and get parsed as):

- ``nCoV-2019_1_LEFT`` (forward primer of amplicon 1)
- ``400_2_out_R`` (reverse primer of amplicon 2)
- ``QIAseq_163-2_LEFT`` (forward primer of amplicon 163)

Alternatively, you can specify the amplicon information explicitly through a
dataset that lists the names of primers that together form any given amplicon.
In it, primer names (exactly matching those in the primer info dataset) need to
be TAB-separated with one line per amplicon.

.. class:: Warning mark

   Preprocessing of the BAM input with ivar trim is essential for this tool to
   work because only ``ivar trim`` can add required primer information to the
   BAM auxillary data of every read.

ivar documentation can be found at `<https://andersen-lab.github.io/ivar/html/manualpage.html>`__.
    ]]></help>
    <expand macro="citations" />
</tool>