view variant.xml @ 5:19f3b583a9b3 draft

"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/medaka commit 13769e7d51b30a1d15eb62a9ba89ee2064f3ddc3"

<?xml version="1.0"?>
<tool id="medaka_variant" name="medaka variant tool" version="@TOOL_VERSION@+galaxy3" profile="@PROFILE@">
    <description>Probability decoding</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements">
        <requirement type="package" version="1.4.1">scipy</requirement>
    </expand>

    <expand macro="version_command"/>
    <command detect_errors="exit_code"><![CDATA[
## initialize
@REF_FASTA@


#if $pool.pool_mode == "Yes":
    ## run
    medaka variant
    ## optional
    --debug
    #if $regions
        --regions '${regions}'
    #end if
    $verbose
    ## required
    reference.fa
    #for $current in $pool.inputs
        '$current'
    #end for
    '$out_result' ## output

    2>&1 | tee '$out_log'
#elif $pool.pool_mode == "No":
    ## run
    medaka variant
    ## optional
    --debug
    #if $regions
        --regions '${regions}'
    #end if
    $verbose
    ## required
    reference.fa
    '$pool.input'
    '$out_result' ##output
    2>&1 | tee '$out_log'
#end if

#if $out_annotated
    && samtools faidx reference.fa
    && grep -v "^#" '$out_result' > tmp.txt
    && cut -f 1,2 tmp.txt > positions.txt
    && samtools mpileup -d 0 --positions positions.txt -Q $output_annotated.min_basequal -q $output_annotated.min_mapqual -f reference.fa '$output_annotated.in_bam' > mpileup.txt
    && python '$__tool_directory__/annotateVCF.py' '$out_result' mpileup.txt '$out_annotated'
#end if
    ]]></command>
    <inputs>
        <conditional name="pool">
            <param name="pool_mode"  type="select" label="Are you pooling HDF5 datasets?">
                <option value="No" selected="True">No</option>
                <option value="Yes">Yes</option>
            </param>
            <when value="Yes">
                <param name="inputs" type="data" format="h5" multiple="true" label="Select consensus file(s)"/>
            </when>
            <when value="No">
                <param name="input" type="data" format="h5" label="Select consensus file(s)"/>
            </when>
        </conditional>
        <expand macro="reference"/>
        <param argument="--regions" type="text" value="" optional="true" label="Set reference names to limit variant calling" help="Separated by ','.">
            <sanitizer invalid_char="">
                <valid initial="string.ascii_letters,string.digits">
                    <add value="_"/>
                    <add value=","/>
                    <add value="."/>
                </valid>
            </sanitizer>
        </param>
        <param argument="--verbose" type="boolean" truevalue="--verbose" falsevalue="" label="Populate VCF info fields?"/>
        <conditional name="output_annotated">
            <param name="output_annotated_select" type="select" label="Output annotated VCF?" help="Annotate allele frequency, depth of coverage, etc for each variant (requires BAM file)">
                <option value="true" selected="true">Output annotated VCF</option>
                <option value="false">Don't output annotated VCF</option>
            </param>
            <when value="true">
                <param name="in_bam" type="data" format="bam" optional="false" label="BAM to annotate the VCF" />
                <param name="min_basequal" type="integer" value="0" min="0" max="50" label="Minimum base quality" help="The minimum base quality (default: 0) at a given variant position required to include a read in the calculation of variant statistics (samtools mpileup -Q)" />
                <param name="min_mapqual" type="integer" value="0" min="0" max="60" label="Minimum mapping quality" help="The minimum mapping quality (default: 0) required for a read to be considered in calculation of variant statistics (samtools mpileup -q)" />
            </when>
            <when value="false"/>
        </conditional>
        <param name="output_log_bool" type="boolean" label="Output log file?" checked="true"/>
    </inputs>
    <outputs>
        <!-- standard -->
        <data name="out_result" format="vcf" label="${tool.name} on ${on_string}: Result"/>
        <!-- optional -->
        <data name="out_annotated" format="vcf" label="${tool.name} on ${on_string}: Annotated">
            <filter>output_annotated_select</filter>
        </data>
        <data name="out_log" format="tabular" label="${tool.name} on ${on_string}: Log">
            <filter>output_log_bool</filter>
        </data>
    </outputs>
    <tests>
        <!-- #1 default -->
        <test expect_num_outputs="3">
            <conditional name="pool">
                <param name="pool_mode" value="Yes"/>
                <param name="inputs" value="medaka_test.hdf,medaka_test.hdf"/>
            </conditional>
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history"/>
                <param name="ref_file" value="ref.fasta"/>
            </conditional>
            <conditional name="output_annotated">
                <param name="output_annotated_select" value="true"/>
                <param name="in_bam" value="medaka_test.bam"/>
            </conditional>
            <param name="output_log_bool" value="true"/>
            
            <output name="out_result">
                <assert_contents>
                    <has_n_lines n="9"/>
                    <has_line line="##fileformat=VCFv4.1" />
                    <has_line line="##medaka_version=1.0.3" />
                </assert_contents>
            </output>
            <output name="out_annotated">
                <assert_contents>
                    <has_n_lines n="16"/>
                    <has_line line="##fileformat=VCFv4.1" />
                    <has_line line="##medaka_version=1.0.3" />
                    <has_line line="##annotateVCFVersion=0.2" />
                </assert_contents>
            </output>
            <output name="out_log">
                <assert_contents>
                    <has_n_lines n="10" />
                </assert_contents>
            </output>
        </test>
        <!--No pooling-->
        <test expect_num_outputs="3">
            <conditional name="pool">
                <param name="pool_mode" value="No"/>
                <param name="input" value="medaka_test.hdf"/>
            </conditional>
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history"/>
                <param name="ref_file" value="ref.fasta"/>
            </conditional>
            <conditional name="output_annotated">
                <param name="output_annotated_select" value="true"/>
                <param name="in_bam" value="medaka_test.bam"/>
            </conditional>
            <param name="output_log_bool" value="true"/>
            
            <output name="out_result">
                <assert_contents>
                    <has_n_lines n="9"/>
                    <has_line line="##fileformat=VCFv4.1" />
                    <has_line line="##medaka_version=1.0.3" />
                </assert_contents>
            </output>
            <output name="out_annotated">
                <assert_contents>
                    <has_n_lines n="16"/>
                    <has_line line="##fileformat=VCFv4.1" />
                    <has_line line="##medaka_version=1.0.3" />
                    <has_line line="##annotateVCFVersion=0.2" />
                </assert_contents>
            </output>
            <output name="out_log">
                <assert_contents>
                    <has_n_lines n="8" />
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[
.. class:: infomark

**What it does**

@WID@

The module *variant* decodes probabilities.

**Input**

- reference sequence (FASTA)
- (several) consensus files (H5/HDF)

**Output**

- decoded probabilities (VCF)

**References**

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