comparison hicBuildMatrix.xml @ 2:eb0d14cdf8b8 draft

planemo upload for repository https://github.com/maxplanck-ie/HiCExplorer/tree/master/galaxy/wrapper/ commit 4d61b6bf2fed275ab38c226d0c4390b095a38251

1:3d56e784d0bd

    <macros>
        <token name="@BINARY@">hicBuildMatrix</token>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements" >
        <requirement type="package" version="1.3.1">samtools</requirement>
    </expand>
    <command>
<![CDATA[

        mkdir ./QCfolder &&
        mkdir $qc.files_path
        &&

        hicBuildMatrix

            --samFiles
            #for $repeat in $samFiles:
                '${repeat.samFile}'
            #end for

            #if $binSize:
                --binSize $binSize
            #end if

            #if $restrictionCutFile:
                --restrictionCutFile '$restrictionCutFile'
            #end if

            #if $minDistance:
                --minDistance $minDistance
            #end if

            #if $maxDistance:
                --maxDistance $maxDistance
            #end if

            #if $restrictionSequence:
                --restrictionSequence '$restrictionSequence'
            #end if

            #end if

            --outFileName ./matrix.h5
            --outBam ./unsorted.bam

            #if $restrictionCutFile:
                --restrictionCutFile '$restrictionCutFile'
            #end if

            $removeSelfCircles

            #if $minMappingQuality and $minMappingQuality is not None:
                --minMappingQuality $minMappingQuality
            #end if

            --QCfolder ./QCfolder
        &&
        mv ./QCfolder/* $qc.files_path/
        &&
        mv $qc.files_path/hicQC.html $qc
        &&
        samtools sort ./unsorted.bam -o sorted.bam

]]>
    </command>
    <inputs>
        <repeat max="2" min="2" name="samFiles" title="Sam/Bam files to process">
            <param name="samFile" type="data" format="sam,bam"/>
        </repeat>

        <param argument="--restrictionCutFile" type="data" format="bed" optional="True"
            help="BED file with all restriction cut places. Should contain only mappable restriction sites. If given, the
                  bins are set to match the restriction fragments (i.e. the region between one restriction site and the next)"/>

        <param name="binSize" type="integer" value="" optional="true" label="Bin size in bp"
            help="If used, the restriction cut places (if given) are used to only consider reads that are in the vicinity of the resctriction sites.
            Otherwise all reads in the interval are considered. " argument="--binSize"/>
        
        <param name="restrictionCutFile" type="data" format="bed" optional="true" label="BED file with all restriction cut places"
            help="Should contaion only  mappable restriction sites. If given, the bins are set to match the restriction fragments
            (i.e. the region between one restriction site and the next)." argument="--restrictionCutFile"/>

        <param name="minDistance" type="integer" value="" optional="true" label="Minimum distance between restriction sites"
            help="Restriction sites that are closer that this distance are merged into one.
            This option only applies if --restrictionCutFile is given." argument="--minDistance"/>

        <param name="maxDistance" type="integer" value="" optional="true" label="Maximum distance in bp from restriction site to read, to consider a read a valid one"
            help="This option only applies if --restrictionCutFile is given."  argument="--maxDistance"/>

        <param name="restrictionSequence" type="text" optional="true" label="Sequence of the restriction site"
            help="This is used to discard reads that end/start with such sequence and that are considered un-ligated fragments or
            &quot;dangling-ends&quot;. If not given, such statistics will not be available." argument="--restrictionSequence"/>

        <expand macro="region" />

        <param argument="--removeSelfCircles" type="boolean" truevalue="--removeSelfCircles" falsevalue=""
            label="Save the matrix of values underlying the heatmap" help="If set, outward facing reads, at a distance of less thatn 25kbs are removed." />

        <expand macro="minMappingQuality" />

    </inputs>
    <outputs>
        <data name="outBam" from_work_dir="sorted.bam" format="bam" label="${tool.name} BAM file on ${on_string}"/>
        <data name="outFileName" from_work_dir="matrix.h5" format="h5" label="${tool.name} MATRIX on ${on_string}"/>
        <data name="qc" format="html" label="${tool.name} QC"/>

                <param name="samFile" value="small_test_R1_unsorted.bam" ftype="sam" />
            </repeat>
            <repeat name="samFiles">
                <param name="samFile" value="small_test_R2_unsorted.bam" ftype="sam"/>
            </repeat>
            <param name="binSize" value="5000"/>
            <output name="outBam" file="hicBuildMatrix_result1.bam" ftype="bam"/>
            <output name="outFileName" file="hicBuildMatrix_result1.h5" ftype="h5" compare="sim_size"/>
        </test>
    </tests>
    <help><![CDATA[

**What it does**

2:eb0d14cdf8b8

    <macros>
        <token name="@BINARY@">hicBuildMatrix</token>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements" >
        <requirement type="package" version="1.6">samtools</requirement>
    </expand>
    <command detect_errors="exit_code"><![CDATA[

        mkdir ./QCfolder &&
        mkdir $qc.files_path &&
        hicBuildMatrix

            --samFiles
            #for $repeat in $samFiles:
                '${repeat.samFile}'
            #end for

            #if $restrictionCutFileBinSize_conditional.restrictionCutFileBinSize_selector == "optionRestrictionCutFile":
                --restrictionCutFile '$restrictionCutFileBinSize_conditional.restrictionCutFile'
                --minDistance $restrictionCutFileBinSize_conditional.minDistance
                --maxDistance $restrictionCutFileBinSize_conditional.maxDistance          
            #end if

            #if $restrictionCutFileBinSize_conditional.restrictionCutFileBinSize_selector == "optionBinSize":
                --binSize $restrictionCutFileBinSize_conditional.binSize
            #end if


            #if $restrictionSequence:
                --restrictionSequence '$restrictionSequence'
            #end if

            #end if

            --outFileName ./matrix.h5
            --outBam ./unsorted.bam

            $keepSelfCircles

            #if $minMappingQuality and $minMappingQuality is not None:
                --minMappingQuality $minMappingQuality
            #end if

            #if $danglingSequence:
                --danglingSequence '$danglingSequence'
            #end if

            --threads @THREADS@
            
            --QCfolder ./QCfolder
        &&
        mv ./QCfolder/* $qc.files_path/
        &&
        mv $qc.files_path/hicQC.html $qc
        && samtools sort ./unsorted.bam -o sorted.bam
       
]]>
    </command>
    <inputs>
        <repeat max="2" min="2" name="samFiles" title="Sam/Bam files to process">
            <param name="samFile" type="data" format="sam,bam"/>
        </repeat>
        <conditional name="restrictionCutFileBinSize_conditional">
            <param name="restrictionCutFileBinSize_selector" type="select" label="Choose to use a restriction cut file or a bin size">
                <option value="optionRestrictionCutFile">Restriction cut file</option>
                <option value="optionBinSize" selected="True">Bin size</option>
            </param>
            <when value="optionRestrictionCutFile">
                <param argument="--restrictionCutFile" type="data" format="bed" optional="True"
                        help="BED file with all restriction cut places. Should contain only mappable restriction sites. If given, the
                            bins are set to match the restriction fragments (i.e. the region between one restriction site and the next)"/>
                <param name="restrictionCutFile" type="data" format="bed" optional="true" label="BED file with all restriction cut places"
                        help="Should contaion only  mappable restriction sites. If given, the bins are set to match the restriction fragments
                        (i.e. the region between one restriction site and the next)." argument="--restrictionCutFile"/>
                <param name="minDistance" type="integer" value="" optional="true" label="Minimum distance between restriction sites"
                        help="Restriction sites that are closer that this distance are merged into one.
                        This option only applies if --restrictionCutFile is given." argument="--minDistance"/>

                <param name="maxDistance" type="integer" value="" optional="true" label="Maximum distance in bp from restriction site to read, to consider a read a valid one"
                        help="This option only applies if --restrictionCutFile is given."  argument="--maxDistance"/>
            </when>
            <when value="optionBinSize">
                <param name="binSize" type="integer" value="" optional="true" label="Bin size in bp"
                        help="If used, the restriction cut places (if given) are used to only consider reads that are in the vicinity of the resctriction sites.
                        Otherwise all reads in the interval are considered. " argument="--binSize"/>
            </when>
        </conditional>

        <param name="restrictionSequence" type="text" optional="true" label="Sequence of the restriction site"
            help="This is used to discard reads that end/start with such sequence and that are considered un-ligated fragments or
            &quot;dangling-ends&quot;. If not given, such statistics will not be available." argument="--restrictionSequence"/>

        <expand macro="region" />

        <param argument="--keepSelfCircles" type="boolean" truevalue="--keepSelfCircles" falsevalue=""
            label="Keep self circles" help="If set, outward facing reads without any restriction fragment (self circles) are kept. '
 +                             'They will be counted and shown in the QC plots." />

        <expand macro="minMappingQuality" />

        <param name="danglingSequence" type="text" optional="true" label="The dangling sequence"
            help="Dangling end sequence left by the restriction enzyme. For DpnII for example, 
                    the dangling end is the same restriction sequence. This is used
                    to discard reads that end/start with such sequence 
                    and that are considered un-ligated fragments or 
                    'dangling-ends'. If not given, such statistics will 
                    not be available." argument="--danglingSequence"/>
    </inputs>
    <outputs>
        <data name="outBam" from_work_dir="sorted.bam" format="bam" label="${tool.name} BAM file on ${on_string}"/>
        <data name="outFileName" from_work_dir="matrix.h5" format="h5" label="${tool.name} MATRIX on ${on_string}"/>
        <data name="qc" format="html" label="${tool.name} QC"/>

                <param name="samFile" value="small_test_R1_unsorted.bam" ftype="sam" />
            </repeat>
            <repeat name="samFiles">
                <param name="samFile" value="small_test_R2_unsorted.bam" ftype="sam"/>
            </repeat>
            <conditional name="restrictionCutFileBinSize_conditional">
                <param name="restrictionCutFileBinSize_selector" value="optionBinSize"/>
                <param name="binSize" value="5000"/>
            </conditional>
            <output name="outBam" file="small_test_matrix_result_sorted.bam" ftype="bam"/>
            <output name="outFileName" file="small_test_matrix.h5" ftype="h5" compare="sim_size" delta="30000"/>
        </test>
    </tests>
    <help><![CDATA[

**What it does**