annotate data_manager/macros.xml @ 5:f5eb9afa8f8a draft

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/data_managers/data_manager_star_index_builder commit 9b68f6ae375aed38493f8399b8572347c750336d

<macros>
    <!-- REMEMBER to bump the version of rna_star_index_builder_data_manager
    whenever you make changes to the following two version tokens!
    The data manager uses a symlink to this macro file to keep the versions in
    sync. -->
    <!-- STAR version to be used -->
    <token name="@VERSION@">2.7.2a</token>
    <!-- STAR index version compatible with this version of STAR
    This is the STAR version that introduced the index structure expected
    by the current version.
    It can be found for any specific version of STAR with:
    STAR -h | grep versionGenome
    or by looking for the versionGenome parameter in source/parametersDefault
    of STAR's source code -->
    <token name="@IDX_VERSION@">2.7.1a</token>

    <xml name="requirements">
        <requirements>
            <requirement type="package" version="@VERSION@">star</requirement>
            <requirement type="package" version="1.9">samtools</requirement>
        </requirements>
    </xml>

    <xml name="index_selection" token_with_gene_model="1">
        <param argument="--genomeDir" name="genomeDir" type="select"
        label="Select reference genome"
        help="If your genome of interest is not listed, contact the Galaxy team">
            <options from_data_table="rnastar_index2_versioned">
                <filter type="static_value" column="4" value="@WITH_GENE_MODEL@" />
                <filter type="static_value" column="5" value="@IDX_VERSION@" />
                <filter type="sort_by" column="2" />
                <validator type="no_options" message="No indexes are available for the selected input dataset" />
            </options>
        </param>
    </xml>

    <token name="@FASTQ_GZ_OPTION@">
        --readFilesCommand zcat
    </token>
    <xml name="citations">
        <citations>
            <citation type="doi">10.1093/bioinformatics/bts635</citation>
        </citations>
    </xml>
    <xml name="@SJDBOPTIONS@" token_optional="true">
         <param argument="--sjdbGTFfile" type="data" format="gff3,gtf" label="Gene model (gff3,gtf) file for splice junctions" optional="@OPTIONAL@" help="Exon junction information for mapping splices"/>
         <param argument="--sjdbOverhang" type="integer" min="1" value="100" label="Length of the genomic sequence around annotated junctions" help="Used in constructing the splice junctions database. Ideal value is ReadLength-1"/>
    </xml>
    <xml name="dbKeyActions">
        <actions>
            <conditional name="refGenomeSource.geneSource">
                <when value="indexed">
                    <action type="metadata" name="dbkey">
                        <option type="from_data_table" name="rnastar_index2_versioned" column="1" offset="0">
                            <filter type="param_value" column="0" value="#" compare="startswith" keep="False"/>
                            <filter type="param_value" ref="refGenomeSource.GTFconditional.genomeDir" column="0"/>
                        </option>
                    </action>
                </when>
                <when value="history">
                    <action type="metadata" name="dbkey">
                        <option type="from_param" name="refGenomeSource.genomeFastaFiles" param_attribute="dbkey" />
                    </action>
                </when>
            </conditional>
        </actions>
    </xml>
    <token name="@TEMPINDEX@"><![CDATA[
    ## Create temporary index for custom reference
    #if str($refGenomeSource.geneSource) == 'history':
        mkdir -p tempstargenomedir &&
        STAR
            --runMode genomeGenerate
            --genomeDir 'tempstargenomedir'
            --genomeFastaFiles '${refGenomeSource.genomeFastaFiles}'
            ## Handle difference between indices with/without annotations
            #if str($refGenomeSource.GTFconditional.GTFselect) == 'with-gtf':
                --sjdbOverhang '${refGenomeSource.GTFconditional.sjdbOverhang}'
                --sjdbGTFfile '${refGenomeSource.GTFconditional.sjdbGTFfile}'
                #if str($refGenomeSource.GTFconditional.sjdbGTFfile.ext) == 'gff3':
                    --sjdbGTFtagExonParentTranscript Parent
                #end if
            #end if
            #if str($refGenomeSource.genomeSAindexNbases):
                --genomeSAindexNbases ${refGenomeSource.genomeSAindexNbases}
            #end if
            --runThreadN \${GALAXY_SLOTS:-4}
        &&
    #end if
    ]]></token>
    <token name="@REFGENOMEHANDLING" ><![CDATA[
    --runThreadN \${GALAXY_SLOTS:-4}
    --genomeLoad NoSharedMemory
    --genomeDir
    #if str($refGenomeSource.geneSource) == 'history':
        tempstargenomedir
    #else:
        '${refGenomeSource.GTFconditional.genomeDir.fields.path}'
        ## Handle difference between indices with/without annotations
        #if str($refGenomeSource.GTFconditional.GTFselect) == 'with-gtf':
            #if $refGenomeSource.GTFconditional.sjdbGTFfile:
                --sjdbOverhang $refGenomeSource.GTFconditional.sjdbOverhang
                --sjdbGTFfile '${refGenomeSource.GTFconditional.sjdbGTFfile}'
                #if str($refGenomeSource.GTFconditional.sjdbGTFfile.ext) == 'gff3':
                    --sjdbGTFtagExonParentTranscript Parent
                #end if
            #end if
        #end if
        #end if
        ]]></token>
    <xml name="stdio" >
        <stdio>
            <regex match="FATAL error" source="both" level="fatal"/>
            <regex match="EXITING: FATAL INPUT ERROR:" source="both" level="fatal"/>
            <regex match="EXITING: fatal error trying to allocate genome arrays, exception thrown: std::bad_alloc" source="both" level="fatal"/>
            <regex match="\[sam_read1\] missing header\? Abort!" source="both" level="fatal"/>
        </stdio>
    </xml>
    <xml name="refgenomehandling" >
        <conditional name="refGenomeSource">
            <param name="geneSource" type="select" label="Custom or built-in reference genome" help="Built-ins were indexed using default options">
                <option value="indexed" selected="true">Use a built-in index</option>
                <option value="history">Use reference genome from history and create temporary index</option>
            </param>
            <when value="indexed">
                <conditional name="GTFconditional">
                    <param name="GTFselect" type="select"
                           label="Reference genome with or without an annotation"
                           help="Select the '... with builtin gene-model' option to select from the list of available indexes that were built with splice junction information. Select the '... without builtin gene-model' option to select from the list of available indexes without annotated splice junctions.">
                        <option value="without-gtf">use genome reference with builtin gene-model</option>
                        <option value="with-gtf">use genome reference without builtin gene-model</option>
                    </param>
                    <when value="with-gtf">
                        <expand macro="index_selection" with_gene_model="0" />
                        <expand macro="@SJDBOPTIONS@" />
                    </when>
                    <when value="without-gtf">
                        <expand macro="index_selection" with_gene_model="1" />
                    </when>
                </conditional>
            </when>
            <when value="history">
                <param argument="--genomeFastaFiles" type="data" format="fasta" label="Select a reference genome" />
                <!-- Currently, this parameter is not exposed in the wrapper,
                     but used only in the tests to avoid excessive index sizes for
                     the tiny test genomes. -->
                <param name="genomeSAindexNbases" type="hidden" value="" />
                <conditional name="GTFconditional">
                    <param name="GTFselect" type="select"
                           label="Build index with our without known splice junctions annotation"
                           help="To build an index with known splice junctions annotated, you will have to provide a GTF or GFF3 dataset that describes the gene models (the location of genes, transcripts and exons) known for the reference genome.">
                        <option value="without-gtf">build index without gene-model</option>
                        <option value="with-gtf">build index with gene-model</option>
                    </param>
                    <when value="with-gtf">
                        <expand macro="@SJDBOPTIONS@" optional="false"/>
                    </when>
                    <when value="without-gtf" />
                </conditional>
            </when>
        </conditional>
    </xml>
</macros>