<tool id="bismark_deduplicate" name="Bismark Deduplicate" version="0.20.0" profile="18.01">
    <description>Deduplicates reads mapped by Bismark</description>
    <requirements>
        <requirement type="package" version="0.20.0">bismark</requirement>
        <requirement type="package" version="1.8">samtools</requirement>
        <requirement type="package" version="2.3.4.2">bowtie2</requirement>
    </requirements>
    <command><![CDATA[
        python '$__tool_directory__/bismark_deduplicate_wrapper.py'

        --single_or_paired $sPaired

        --input '$mapping_output'
        --output_report '$output_report'
        --output_bam '$output_bam'
        #if $separate_logfile:
           --log_report '$log_report'
        #end if
]]>
    </command>
    <inputs>
        <param name="sPaired" type="select" label="Is this library mate-paired?">
            <option value="single">Single-end</option>
            <option value="paired">Paired-end</option>
        </param>
        <param name="mapping_output" type="data" format="qname_input_sorted.bam,bam"
               label="Submit the resulting bam/sam file from Bismark bisulfite mapper"/>
        <param name="separate_logfile" type="boolean" truevalue="true" falsevalue="false" checked="False"
               label="Create a separate logfile, otherwise logs are added to the dataset info."/>
    </inputs>

    <outputs>
        <data name="output_bam" format="qname_sorted.bam"
              label="${tool.name} on ${on_string}: deduplicated mapped reads"/>
        <data name="output_report" format="txt" label="${tool.name} on ${on_string}: deduplication report"/>
        <data name="log_report" format="txt" label="${tool.name} on ${on_string}: log report (tool stdout)">
            <filter>( separate_logfile is True )</filter>
        </data>
    </outputs>

    <tests>
        <test>
            <param name="sPaired" value="single"/>
            <param name="mapping_output" value="mapped_reads.bam" ftype="qname_sorted.bam"/>
            <output name="output_bam" file="dedup_reads.bam" ftype="qname_sorted.bam"/>
            <output name="output_report" file="dedup_report.txt" ftype="txt"/>
        </test>
    </tests>

    <help>
        <![CDATA[
**What it does**

	| This tool is supposed to remove alignments to the same position in the genome from the Bismark mapping output (both single and paired-end SAM files), which can arise by e.g. excessive PCR amplification. If sequences align to the same genomic position but on different strands they will be scored individually.
	|
	| Note that deduplication is not recommended for RRBS-type experiments!
	|
	| For single-end alignments only use the start-coordinate of a read will be used for deduplication.
	| For paired-end alignments the start-coordinate of the first read and the end coordinate of the second read will be used for deduplication.

]]>
    </help>
    <citations>
        <citation type="doi">10.1093/bioinformatics/btr167</citation>
    </citations>
</tool>