annotate galaxy/wrappers/FactorFileGenerator.xml @ 1:7a520f7169e1 draft

"planemo upload for repository https://github.com/juliechevalier/GIANT/tree/master commit e2b27d6ff2eab66454f984dbf1a519192f41db97"

<tool name="GIANT-Factor file generator" id="giant_factor_generator" version="0.1.3">
  <description>Generate factor file used by other GIANT tools</description>
  <requirements>
  </requirements>
  <code file="../../src/General_functions.py"/>
  <!--<code file="./src/General_functions.py"/> change for Planemo test-->
  <stdio>
    <regex match="Execution halted"
           source="both"
           level="fatal"
           description="Execution halted, please contact tool developer or administrators." />
    <regex match="Error in"
           source="both"
           level="fatal"
           description="An error occured during R execution, please contact tool developer." />
    <exit_code range="1:9" level="fatal" description="Error during factor file generation, see log file for more information." />
  </stdio>
  <command>	<![CDATA[

  #import imp
  #set $general_functions=$imp.load_source('General_functions', $__tool_directory__+'/../../src/General_functions.py')
  ##change for Planemo test
  ##set $general_functions=$imp.load_source('General_functions', $__tool_directory__+'/src/General_functions.py')

  #if $inputCondition.selection == "CELcollection" and $inputCondition.selectCollectionStrat.how=="group_tags":
    #set $temp_factor_names = list()
    #for $factor in $inputCondition.selectCollectionStrat.factorListBis:
          #set $temp_factor = list()
          #for $level in $factor.valueList:
            #set $temp_level = ''
            #for $group in $level.valueConditions.value:
                #for $file in $inputCondition.inputData.get_datasets_for_group($group):
                    printf "$group\t $file.element_identifier\n" >> ./TAGmatching.csv;
                    #if $temp_level == '':
                      #set $temp_level = str($file.element_identifier)
                    #else:
                      #set $temp_level += ','+str($file.element_identifier)
                    #end if
                #end for
            #end for
            $temp_factor.append( {'valueName':str($level.valueName), 'valueConditions':$temp_level} )
          #end for
          $temp_factor.reverse()
          $temp_factor_names.append( {'factorName':str($factor.factorName), 'valueList':$temp_factor} )
    #end for
  #end if

  #if $inputCondition.selection == "CELcollection":
    #if $inputCondition.selectCollectionStrat.how=="group_tags":
      #set $ret_code=$general_functions.generateFactorFile($inputCondition.inputData.collection,$temp_factor_names,$outputData.file_name,$log.file_name)
    #else:
      #set $ret_code=$general_functions.generateFactorFile($inputCondition.inputData.collection,$inputCondition.selectCollectionStrat.factorListBis,$outputData.file_name,$log.file_name)
    #end if
  #else:
    #set $ret_code=$general_functions.generateFactorFile($inputCondition.inputData,$inputCondition.factorList,$outputData.file_name,$log.file_name)
  #end if

  if [ $ret_code != 0 ]; then
    printf "[ERROR]Error during factor file generation\n" >> $log;
    exit $ret_code;
  fi;

  printf "[INFO]End of tool script" >> $log; 
	]]>
  </command>
  <inputs>
  <param type="text" name="title" value="ConditionsGenerator_toPersonalize" label="Title for output"/>

  <conditional name="inputCondition">
      <param name="selection" type="select" label="Input data type for sample names" force_select="true">
        <option value="normalizedData">Expression tabular file</option>
        <option value="CELfiles">.CEL files</option>
        <option value="CELcollection">.CEL file collection</option>
      </param>
      <when value="normalizedData">
        <param type="data" name="inputData" format="tabular" label="Select a single dataset" optional="false" multiple="false"/>

            <repeat name="factorList" title="Factor">
              <param type="text" name="factorName" value="" label="Factor name"/>
              <repeat name="valueList" title="Value">
                <param type="text" name="valueName" value="" label="Value name"/>
                <param name="valueConditions" type="select" optional="false" multiple="true" label="Select sample sharing this value"
                  refresh_on_change="true"  dynamic_options="get_condition_file_names(inputCondition['inputData'],0)">
                </param>
              </repeat>
            </repeat>

      </when>
      <when value="CELfiles">
        <param type="data" name="inputData" format="cel" label="Select multiple files" optional="false" multiple="true">
        <validator type="empty_field" message="At least two data files should be selected"></validator>
        </param>

            <repeat name="factorList" title="Factor">
              <param type="text" name="factorName" value="" label="Factor name"/>
              <repeat name="valueList" title="Value">
                <param type="text" name="valueName" value="" label="Value name"/>
                <param name="valueConditions" type="select" optional="false" multiple="true" label="Select sample sharing this value"
                  refresh_on_change="true"  dynamic_options="get_condition_file_names(inputCondition['inputData'])">
                </param>
              </repeat>
            </repeat>

      </when>
      <when value="CELcollection">
        <param type="data_collection" name="inputData" format="cel" label="Select a single dataset collection" optional="false" multiple="true">
        <validator type="empty_field" message="One data collection should be selected"></validator>
        </param>

        <conditional name="selectCollectionStrat">
          <param name="how" type="select" label="Select how .CEL files will be assigned to factor values">
              <option value="individualSelection">Select individually each .CEL file from the collection</option>
              <option value="group_tags">Use associated .CEL file tags</option>
          </param>
          <when value="individualSelection">

              <repeat name="factorListBis" title="Factor">
                <param type="text" name="factorName" value="" label="Factor name"/>
                <repeat name="valueList" title="Value">
                  <param type="text" name="valueName" value="" label="Value name"/>
                  <param name="valueConditions" type="select" optional="false" multiple="true" label="Select sample sharing this value"
                    refresh_on_change="true"  dynamic_options="get_condition_file_names(inputCondition['inputData'])">
                  </param>
                </repeat>
              </repeat>

          </when>
          <when value="group_tags">

              <repeat name="factorListBis" title="Factor">
                <param type="text" name="factorName" value="" label="Factor name"/>
                <repeat name="valueList" title="Value">
                  <param type="text" name="valueName" value="" label="Value name"/>
                  <param name="valueConditions" type="group_tag" data_ref="inputData" multiple="true" label="Select groups sharing this value"/>
                </repeat>
              </repeat>

          </when>
        </conditional>
      </when>
  </conditional>
  </inputs>

  <outputs>
    <data format="tabular" name="outputData" label="${title}_conditionsFile"/>
    <data format="txt" name="log" label="${title}_Log" />
  </outputs>
  


 <tests>
  <test maxseconds="3600">
    <conditional name="inputCondition">
      <param name="selection" value="normalizedData" />
      <param name="inputData" value="./NormalizedData.tabular" />
      <repeat name="factorList">
        <param name="factorName" value="Strain" />
        <repeat name="valueList">
          <param name="valueName" value="WT" />
          <param name="valueConditions" value="GSM205769.CEL,GSM205772.CEL,GSM205768.CEL,GSM205767.CEL,GSM205766.CEL,GSM205771.CEL,GSM205770.CEL"/>
        </repeat>
        <repeat name="valueList">
          <param name="valueName" value="KO" />
          <param name="valueConditions" value="GSM205777.CEL,GSM205776.CEL,GSM205781.CEL,GSM205773.CEL,GSM205780.CEL,GSM205779.CEL,GSM205782.CEL,GSM205775.CEL,GSM205774.CEL,GSM205778.CEL"/>
        </repeat>
      </repeat>
      <repeat name="factorList">
        <param name="factorName" value="Treatment" />
        <repeat name="valueList">
          <param name="valueName" value="Control" />
          <param name="valueConditions" value="GSM205777.CEL,GSM205776.CEL,GSM205773.CEL,GSM205775.CEL,GSM205774.CEL,GSM205768.CEL,GSM205767.CEL,GSM205766.CEL"/>
        </repeat>
        <repeat name="valueList">
          <param name="valueName" value="Treat" />
          <param name="valueConditions" value="GSM205781.CEL,GSM205769.CEL,GSM205772.CEL,GSM205780.CEL,GSM205779.CEL,GSM205782.CEL,GSM205778.CEL,GSM205771.CEL,GSM205770.CEL"/>
      </repeat>
      </repeat>
    </conditional>
    <output name="outputData" file="./FactorFileGenerator/output/conditionsFile.csv" />
    <output name="log" file="./FactorFileGenerator/output/outputLog.txt" />    
  </test>
 </tests> 



  <help>
<![CDATA[
**What it does ?**

This tool generates factor information file used by other tools of GIANT tool suite.

-----

**Parameters**

\- **Title** to personalize output file names (please avoid special characters and spaces).

\- **Input Data** used only to extract sample names

- **Expression tabular file** with samples as columns and genes as rows (only the header row will be used to extract sample names).

    ::

        Conditions  157_(HuGene-2_0-st).CEL 156_(HuGene-2_0-st).CEL  155_(HuGene-2_0-st).CEL    154_(HuGene-2_0-st).CEL                        
        DDX11L2     4.500872                4.429759                 4.780281                   4.996189             
        MIR1302-2   3.415065                3.520472                 3.471503                   3.567988           
        OR4F5       3.737956                3.011586                 3.424494                   3.497545
        VWA1        5.189621                5.129595                 4.806793                   5.227014

OR

- **.CEL files** of your study (you should select multiple .CEL files).

OR

- **.CEL file collection** of your study (you should select a unique collection file).

  \- **Individual selection** of files to associate to factor values.
  
    or

  \- **Tag selection** to associate samples sharing the same tag to factor values.

\- **Factor definition**

- **Factor name** to discriminate between samples as 'Treatments', 'Year', 'Strain' (please avoid special characters)

- **Value name** of different states for the current factor as 'KO' or 'WT' for 'Strain' factor (please avoid special characters)

- **Select sample/tag** to assign to current value

-----

**Outputs**

- **Factor information tabular file**  with factors as columns and samples as rows (header row contains factor names and first column sample names).

    ::

        Conditions                Sex   Treatment Reaction
        138_(HuGene-2_0-st).CEL   1     TreatA    Pos
        148_(HuGene-2_0-st).CEL   0     NoTreat   Pos
        139_(HuGene-2_0-st).CEL   0     TreatB    Neg
        149_(HuGene-2_0-st).CEL   0     NoTreat   Neg

- **LOG file** for job log. If you see errors, please attached this in the bug report

]]>  </help>

</tool>