<tool id="gatk_validate_variants" name="Validate Variants" version="0.0.4">
  <description></description>
  <requirements>
      <requirement type="package" version="1.4">gatk</requirement>
  </requirements>
  <macros>
    <import>gatk_macros.xml</import>
  </macros>
  <command interpreter="python">gatk_wrapper.py
   --max_jvm_heap_fraction "1"
   --stdout "${output_log}"
   -d "--variant:variant,%(file_type)s" "${reference_source.input_variant}" "${reference_source.input_variant.ext}" "input_variant"
   -p 'java 
    -jar "\$JAVA_JAR_PATH/GenomeAnalysisTK.jar"
    -T "ValidateVariants"
    
    -et "NO_ET" ##ET no phone home
    ##--num_threads 4 ##hard coded, for now
    ##-log "${output_log}" ##don't use this to log to file, instead directly capture stdout
    #if $reference_source.reference_source_selector != "history":
        -R "${reference_source.ref_file.fields.path}"
    #end if
    ${warn_on_errors}
    ${do_not_validate_filtered_records}
   '
   
    #if str( $dbsnp_rod_bind_type.dbsnp_rod_bind_type_selector ) == 'set_dbsnp':
        -d "--dbsnp:${dbsnp_rod_bind_type.dbsnp_rod_name},%(file_type)s" "${dbsnp_rod_bind_type.dbsnp_input_rod}" "${dbsnp_rod_bind_type.dbsnp_input_rod.ext}" "input_dbsnp_${dbsnp_rod_bind_type.dbsnp_rod_name}"
    #end if
   
    #include source=$standard_gatk_options#    
  </command>
  <inputs>
    
    <conditional name="reference_source">
      <expand macro="reference_source_selector_param" />
      <when value="cached">
        <param name="input_variant" type="data" format="vcf" label="Input variant file" help="-V,--variant &amp;lt;variant&amp;gt;" />
        <param name="ref_file" type="select" label="Using reference genome" help="-R,--reference_sequence &amp;lt;reference_sequence&amp;gt;">
          <options from_data_table="gatk_picard_indexes">
            <filter type="data_meta" key="dbkey" ref="input_variant" column="dbkey"/>
          </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="input_variant" type="data" format="vcf" label="Input variant file" help="-V,--variant &amp;lt;variant&amp;gt;" />
        <param name="ref_file" type="data" format="fasta" label="Using reference file" help="-R,--reference_sequence &amp;lt;reference_sequence&amp;gt;" />
      </when>
    </conditional>
    
    <conditional name="dbsnp_rod_bind_type">
      <param name="dbsnp_rod_bind_type_selector" type="select" label="Provide a dbSNP reference-ordered data file" help="-D,--dbsnp &amp;lt;dbsnp&amp;gt;">
        <option value="set_dbsnp" selected="True">Set dbSNP</option>
        <option value="exclude_dbsnp">Don't set dbSNP</option>
      </param>
      <when value="exclude_dbsnp">
        <!-- Do nothing here -->
      </when>
      <when value="set_dbsnp">
        <param name="dbsnp_input_rod" type="data" format="vcf" label="ROD file" />
        <param name="dbsnp_rod_name" type="hidden" value="dbsnp" label="ROD Name"/>
      </when>
    </conditional>
    
    <param name="warn_on_errors" type="boolean" checked="False" truevalue="-warnOnErrors" falsevalue="" label="instead of terminating the run at the first error, print warning messages for each error seen." help="-warnOnErrors,--warnOnErrors"/>
    <param name="do_not_validate_filtered_records" type="boolean" checked="False" truevalue="-doNotValidateFilteredRecords" falsevalue="" label="do not try to validate records that are FILTERed." help="-doNotValidateFilteredRecords,--doNotValidateFilteredRecords"/>
    
    <expand macro="gatk_param_type_conditional" />
    
  </inputs>
  <outputs>
    <data format="txt" name="output_log" label="${tool.name} on ${on_string} (log)" />
  </outputs>
  <tests>
      <test>
          <param name="reference_source_selector" value="history" />
          <param name="ref_file" value="phiX.fasta" ftype="fasta" />
          <param name="input_variant" value="gatk/gatk_variant_annotator/gatk_variant_annotator_out_1.vcf" ftype="vcf" />
          <param name="dbsnp_rod_bind_type_selector" value="set_dbsnp" />
          <param name="dbsnp_input_rod" value="gatk/fake_phiX_variant_locations.vcf" ftype="vcf" />
          <param name="warn_on_errors" value="True"/>
          <param name="do_not_validate_filtered_records" />
          <param name="gatk_param_type_selector" value="basic" />
          <output name="output_log" file="gatk/gatk_validate_variants/gatk_validate_variants_out_1.log.contains" compare="contains" />
      </test>
  </tests>
  <help>
**What it does**

Validates a variants file.

For more information on using the ValidateVariants module, see this `tool specific page &lt;http://www.broadinstitute.org/gsa/wiki/index.php/VariantValidator&gt;`_.

To learn about best practices for variant detection using GATK, see this `overview &lt;http://www.broadinstitute.org/gsa/wiki/index.php/Best_Practice_Variant_Detection_with_the_GATK_v3&gt;`_.

If you encounter errors, please view the `GATK FAQ &lt;http://www.broadinstitute.org/gsa/wiki/index.php/Frequently_Asked_Questions&gt;`_.

------

**Inputs**

GenomeAnalysisTK: ValidateVariants accepts variant files as input.


**Outputs**

The output is a log of variant validation.


Go `here &lt;http://www.broadinstitute.org/gsa/wiki/index.php/Input_files_for_the_GATK&gt;`_ for details on GATK file formats.

-------

**Settings**::

 doNotValidateFilteredRecords    should we skip validation on filtered records?
 warnOnErrors                    should we just emit warnings on errors instead of terminating the run?

@CITATION_SECTION@
  </help>
</tool>