comparison bamtools.xml @ 0:073a6f5950d0 draft

Uploaded

0:073a6f5950d0

<?xml version="1.0"?>
<tool id="bamtools" name="Convert, Merge, Randomize" version="0.0.1">
  <description>BAM datasets and perform other transformations</description>
  <requirements>
    <requirement type="package" version="2.3.0_2d7685d2ae">bamtools</requirement>
    <requirement type="package" version="0.1.18">samtools</requirement>
  </requirements>
  
  <command>
   ##set up input files
  
  #for $bam_count, $input_bam in enumerate( $input_bams ):
    ln -s "${input_bam.input_bam}" "localbam_${bam_count}.bam" &amp;&amp;
    ln -s "${input_bam.input_bam.metadata.bam_index}" "localbam_${bam_count}.bam.bai" &amp;&amp;
  #end for
   
  #if str( $analysis_type.analysis_type_selector ) == "convert":
    #if str( $analysis_type.format_type.format_type_selector ) == "pileup":
      #set $reference_fasta_filename = "localref.fa"
      #if str( $analysis_type.format_type.reference_source.reference_source_selector ) == "history":
        ln -s "${analysis_type.format_type.reference_source.ref_file}" "${reference_fasta_filename}" &amp;&amp;
        samtools faidx "${reference_fasta_filename}" 2&gt;&amp;1 || echo "Error running samtools faidx for bamtools convert" &gt;&amp;2 &amp;&amp;
      #else:
        #set $reference_fasta_filename = str( $analysis_type.format_type.reference_source.ref_file.fields.path )
      #end if
    #end if
  #end if

    ##finished setting up inputs
    
    ##start bamtools commandline
    
    bamtools
    
    #if str( $analysis_type.analysis_type_selector ) == "convert":
    
    convert
    
    -format ${analysis_type.format_type.format_type_selector}
    
      #if str( $analysis_type.format_type.format_type_selector ) == "pileup":
    
      ${analysis_type.format_type.mapqual}
      -fasta "${reference_fasta_filename}"
    
      #elif str( $analysis_type.format_type.format_type_selector ) == "sam":
    
      ${analysis_type.format_type.noheader}
    
      #end if
      
    -out $out_file1
    
    #elif str( $analysis_type.analysis_type_selector ) == "count":
    
    count
    > $out_file1
    
    #elif str( $analysis_type.analysis_type_selector ) == "coverage":
    
    coverage
    -out $out_file1
    
    #elif str( $analysis_type.analysis_type_selector ) == "header":
    
    header
    > $out_file1
    
    #elif str( $analysis_type.analysis_type_selector ) == "merge":
    
    merge
    -out $out_file1
    
    #elif str( $analysis_type.analysis_type_selector ) == "random":
    
    random
    -n ${analysis_type.count}
    -seed ${analysis_type.seed}
    -out $out_file1
    
    #elif str( $analysis_type.analysis_type_selector ) == "revert":
    
    revert
    ${analysis_type.keepDuplicate}
    ${analysis_type.keepQualities}
    -out $out_file1
    
    #elif str( $analysis_type.analysis_type_selector ) == "sort":
    
    sort
    ${analysis_type.byname}
    -out $out_file1
    
    #end if
    
    #for $bam_count, $input_bam in enumerate( $input_bams ):
        -in "localbam_${bam_count}.bam"
    #end for
    
   
  </command>
  <inputs>
    
    <repeat name="input_bams" title="BAM dataset(s) to filter" min="1">
          <param name="input_bam" type="data" format="bam" label="BAM dataset" />
    </repeat>

    <conditional name="analysis_type">
      <param name="analysis_type_selector" type="select" label="Select BAM manipulation" help="See help below for detailed description of each tool">
        <option value="convert">Convert</option>
        <option value="count">Count</option>
        <option value="coverage">Coverage</option>
        <option value="header">Header</option>
        <option value="merge">Merge</option>
        <option value="random">Random</option>
        <option value="revert">Revert</option>
      <!-- The sort option below is commented out as BAM files in Galaxy are reference sorted by dafault. -->
      <!-- Allowing users for sort files may break donstream functionality. -->
      <!-- To enable sort option simply uncomment the line below: -->
      <!--  <option value="sort">Sort</option> -->
      </param>
      <when value="convert">
        <conditional name="format_type">
          <param name="format_type_selector" type="select" help="Select what to convert your BAM to">
           <option value="bed">BED</option>
           <option value="fasta">FASTA</option>
           <option value="fastq">FASTQ</option>
           <option value="json">JSON</option>
           <option value="pileup">Pileup</option>
           <option value="sam">SAM</option>
           <option value="yaml">YAML</option>
          </param>
          <when value="pileup">
            <conditional name="reference_source">
              <param name="reference_source_selector" type="select" label="Choose the source for the reference list">
                <option value="cached">Locally cached</option>
                <option value="history">History</option>
              </param>
              <when value="cached">
                <param name="ref_file" type="select" label="Using reference genome">
                  <options from_data_table="sam_fa_indexes">
                  <!--<filter type="data_meta" key="dbkey" ref="input_bam" column="value"/>-->
                  </options>
                  <validator type="no_options" message="A built-in reference genome is not available for the build associated with the selected input file"/>
                </param>
              </when>
              <when value="history"> <!-- FIX ME!!!! -->
                <param name="ref_file" type="data" format="fasta" label="Using reference file" />
              </when>
            </conditional>
            <param name="mapqual" type="boolean" truevalue="-mapqual" falsevalue="" label="Print quality scores?" />
          </when>
          <when value="sam">
            <param name="noheader" type="boolean" truevalue="-noheader" falsevalue="" label="Do not print header" />
          </when>
        </conditional>
      </when>
      <when value="count">
      <!-- Nothing to be done with count -> just count alignments in the input bam(s) -->
      </when>
      <when value="coverage">
      <!-- Nothing to be done with count -> just count alignments in the input bam(s) -->
      </when>
      <when value="header">
      <!-- Nothing to be done with count -> just count alignments in the input bam(s) -->
      </when>
      <when value="merge">
      <!-- Nothing to be done with count -> just count alignments in the input bam(s) -->
      </when>
      <when value="random">
        <param name="count" type="integer" value="10000" label="Number of random alignments to grab" help="No duplicate checking is perfomed" />
        <param name="seed" type="integer" value="1024" label="Random number generator seed" help="Use the same seed for reproducible results" />
      </when>
      <when value="revert">
        <param name="keepDuplicate" type="boolean" truevalue="-keepDuplicate" falsevalue="" label="Keep duplicates marked" help="Do not remove duplicate marks" />
        <param name="keepQualities" type="boolean" truevalue="-keepQualities" falsevalue="" label="Keep base qualities" help="Do not replace qualities with contect of OQ tag" />
      </when>
      <when value="sort">
        <param name="byname" type="boolean" truevalue="-byname" falsevalue="" label="Sort by name" help="Checked: sort by name; Unchecked: sort by coordinate"/>
      </when>
    </conditional>
    
    </inputs>
    <outputs>
      <data format="txt" name="out_file1">
        <change_format>
          <when input="analysis_type.format_type.format_type_selector" value="bed" format="bed" />
          <when input="analysis_type.format_type.format_type_selector" value="fasta" format="fasta" />
          <when input="analysis_type.format_type.format_type_selector" value="fastq" format="fastq" />
          <when input="analysis_type.format_type.format_type_selector" value="sam" format="sam" />
          <when input="analysis_type.format_type.format_type_selector" value="pileup" format="pileup" />
          <when input="analysis_type.analysis_type_selector" value="coverage" format="tabular" />
          <when input="analysis_type.analysis_type_selector" value="merge" format="bam" />
          <when input="analysis_type.analysis_type_selector" value="random" format="bam" />
          <when input="analysis_type.analysis_type_selector" value="revert" format="bam" />
          <when input="analysis_type.analysis_type_selector" value="sort" format="bam" />
        </change_format>
      </data>  
    </outputs>
    <tests>
      <test>
        <param name="input_bam" ftype="bam" value="bamtools-input1.bam"/>
        <param name="analysis_type_selector" value="convert"/>
        <param name="format_type_selector" value="pileup"/>
        <param name="reference_source_selector" value="history" />
        <param name="mapqual" value="true" />
        <param name="ref_file" ftype="fasta" value="bamtools-fasta.fa"/>
        <output name="output_bam" file="bamtools-convert-pileup.pu" />
      </test>
      <test>
        <param name="input_bam" ftype="bam" value="bamtools-input1.bam"/>
        <param name="analysis_type_selector" value="count"/>
        <output name="output_bam" file="bamtools-count.tab" />
      </test>
      <test>
        <param name="input_bam" ftype="bam" value="bamtools-input1.bam"/>
        <param name="analysis_type_selector" value="coverage"/>
        <output name="output_bam" file="bamtools-coverage.tab" />
      </test>
      <test>
        <param name="input_bam" ftype="bam" value="bamtools-input1.bam"/>
        <param name="analysis_type_selector" value="header"/>
        <output name="output_bam" file="bamtools-header.txt" />
      </test>
    </tests>
  
  <stdio>
    <exit_code range="1:" />
  </stdio>
  
  <help>

**What is does**

BAMTools is a collection of utilities for manipulation on BAM files. It is based on BAMtools suite of tools by Derek Barnett (https://github.com/pezmaster31/bamtools).
This Galaxy implementation of BAMTools utilities includes seven utilities - Convert, Count, Coverage, Header, Merge, Random, and Revert - decsribed in detail below.

-----

**Convert**

Converts BAM dataset(s) into BED, FASTA, FASTQ, JSON, Pileup, SAM, or YAML formats. Note that the conversion to the pileup format requires providing a reference sequence either
cashed at this Galaxy instance, or provided by you as a FASTA dataset from History.

-----

**Count**

Counts a number of alignments in a BAM dataset(s).

-----

**Coverage**

Prints per-base coverage for a BAM dataset.

-----

**Header**

Prints header from a BAM dataset(s).

------

**Merge**

Merges multiple BAM datasets into a single one. Obviously, you need to select multiple BAMs as input, which is done by pressing the "**Add new BAM dataset(s) to filter**" button.

------

**Random**

Grabs a specified number of random lines from BAM dataset(s).

------

**Revert**

Removes duplicate marks and restores original (non-recalibrated) base qualities.

-----

.. class:: infomark

**More information**

Additional information about BAMtools can be found at https://github.com/pezmaster31/bamtools/wiki

  </help>
  <citations>
    <citation type="doi">10.1093/bioinformatics/btr174</citation>
  </citations>
</tool>