<tool id="rseqc_junction_annotation" name="Junction Annotation" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="@GALAXY_VERSION@">
    <description>compares detected splice junctions to reference gene model</description>
    <macros>
        <import>rseqc_macros.xml</import>
    </macros>
    <expand macro="bio_tools"/>
    <expand macro="requirements">
        <!--
            Required due to conda solver bug: https://github.com/conda/conda/issues/6269
            See: https://github.com/galaxyproject/tools-iuc/pull/1578 for more info
        -->
        <requirement type="package" version="4.2.2">r-base</requirement>
    </expand>
    <expand macro="stdio"/>
    <version_command><![CDATA[junction_annotation.py --version]]></version_command>
    <command><![CDATA[
        @BAM_SAM_INPUTS@
        junction_annotation.py
            --input-file 'input.${extension}'
            --refgene '${refgene}'
            --out-prefix output
            --min-intron ${min_intron}
            --mapq ${mapq}
        2> >(tee -a stats.txt >&2)
        ]]>
    </command>
    <inputs>
        <expand macro="bam_sam_param"/>
        <expand macro="refgene_param"/>
        <expand macro="min_intron_param"/>
        <expand macro="mapq_param"/>
        <expand macro="rscript_output_param"/>
    </inputs>
    <outputs>
        <data format="pdf" name="outputpdf" from_work_dir="output.splice_events.pdf" label="${tool.name} on ${on_string}: splice events (PDF)"/>
        <data format="pdf" name="outputjpdf" from_work_dir="output.splice_junction.pdf" label="${tool.name} on ${on_string}: splice junction (PDF)"/>
        <expand macro="xls_output_data" filename="output.junction.xls" label="${tool.name} on ${on_string}: splice junction (XLS)"/>
        <expand macro="rscript_output_data" filename="output.junction_plot.r" label="${tool.name} on ${on_string}: Rscript"/>
        <data format="txt" name="stats" from_work_dir="stats.txt" label="${tool.name} on ${on_string}: stats"/>
    </outputs>
    <tests>
        <test expect_num_outputs="5">
            <param name="input" value="pairend_strandspecific_51mer_hg19_chr1_1-100000.bam"/>
            <param name="refgene" value="hg19_RefSeq_chr1_1-100000.bed" ftype="bed12"/>
            <param name="rscript_output" value="true"/>
            <output name="outputxls" file="output.junction.xls" ftype="tabular"/>
            <output name="outputr" file="output.junction_plot_r"/>
            <output name="outputpdf" file="output.splice_events.pdf" compare="sim_size"/>
            <output name="outputjpdf" file="output.splice_junction.pdf" compare="sim_size"/>
            <output name="stats" file="output.splice_junction.txt" ftype="txt" lines_diff="2"/>
        </test>
    </tests>
    <help><![CDATA[
junction_annotation.py
++++++++++++++++++++++

For a given alignment file (-i) in BAM or SAM format and a reference gene model (-r) in BED
format, this program will compare detected splice junctions to reference gene model. splicing
annotation is performed in two levels: splice event level and splice junction level.

* splice event: An RNA read, especially long read, can be spliced 2 or more times, each time is called a splicing event; In this sense, 100 spliced reads can produce >= 100 splicing events.
* splice junction: multiple splicing events spanning the same intron can be consolidated into one splicing junction.

All detected junctions can be grouped to 3 exclusive categories:

1. Annotated: The junction is part of the gene model. Both splice sites, 5' splice site
   (5'SS) and 3'splice site (3'SS) can be annotated by reference gene model.
2. complete_novel: Complete new junction. Neither of the two splice sites cannot be annotated by gene model
3. partial_novel: One of the splice site (5'SS or 3'SS) is new, while the other splice site is annotated (known)

Inputs
++++++++++++++

Input BAM/SAM file
    Alignment file in BAM/SAM format.

Reference gene model
    Gene model in BED format.

Minimum intron length (default=50)
    Minimum intron length (bp).


Output
++++++++++++++

1. output.junc.anno.junction.xls:
    - chrom ID
    - start position of junction (coordinate is 0 based)
    - end position of junction (coordinate is 1 based)
    - number of splice events supporting this junction
    - 'annotated', 'complete_novel' or 'partial_novel'.
2. output.anno.junction_plot.r: R script to generate pie chart
3. output.splice_junction.pdf: plot of splice junctions
4. output.splice_events.pdf: plot of splice events

.. image:: $PATH_TO_IMAGES/junction.png
   :height: 400 px
   :width: 850 px
   :scale: 80 %

@ABOUT@

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