root/galaxy-central/tools/regVariation/compute_q_values.xml @ 2

<tool id="compute_q_values" name="Compute q-values" version="1.0.0">
  <description>based on multiple simultaneous tests p-values</description>
  
  <command interpreter="perl">
        compute_q_values.pl $inputFile1 $inputLambda2 $inputPI0_method3 $inputFDR_level4 $inputRobust5 $outputFile1 $outputFile2
  </command>

  <inputs>
        <param format="tabular" name="inputFile1" type="data" label="Select the p-values file"/>
        
        <param type="text" name="inputLambda2" size="100" value="sequence_from_0_to_0.95_increment_0.05" label="What is the lambda value?" help="Either choose the default sequence or one deciaml value between 0 and 1"/>
        
        <param name="inputPI0_method3" type="select" label="Choose the PI method:">
        <option value="smoother">smoother</option>
        <option value="bootstrap">bootstrap</option>
    </param>
    
    <param type="float" name="inputFDR_level4" size="5" value="" label="What is the FDR level?" help="The FDR level must be between 0 and 1"/>
    
    <param name="inputRobust5" type="select" label="Do you want to make the estimate more robust:" help="Choose TRUE for small p-values">
                <option value="FALSE">FALSE</option>
        <option value="TRUE">TRUE</option>
    </param>
  </inputs>
  
  <outputs>
    <data format="tabular" name="outputFile1"/>
    <data format="pdf" name="outputFile2"/>
  </outputs>

  <tests>
        <test>
                <param name="inputFile1" value="p_values.tabular" ftype="tabular" />
        <param name="inputLambda2" value="sequence_from_0_to_0.95_increment_0.05" />
        <param name="inputPI0_method3" value="smoother" />
        <param name="inputFDR_level4" value="0.05" />
        <param name="inputRobust5" value="FALSE" />
        <output name="outputFile1" file="q_values.tabular" />
        <output name="outputFile1" file="p_q_hists_Q_plots.pdf" />
                </test>
  </tests>
        
  <help> 

.. class:: infomark

**What it does**

This program computes the q-values based on the p-values of multiple simultaneous tests. The q-values are computed using a specific R package, created by John Storey and Alan Dabney, called "qvalue". The program takes five inputs:
 
- The first input is a TABULAR format file consisting of one column only that represents the p-values of multiple simultaneous tests, one line for every p-value. 
- The second input is the lambda parameter. The user can choose either the default: seq(0, 0.95, 0.05) or a decimal number between 0.0 and 1.0.
- The third input is PI method which is either "smoother" or "bootstrap".
- The fourth input is the FDR (false discovery rate) level which is a decimal number between 0.0 and 1.0.
- The fifth input is either TRUE or FALSE for the estimate robustness. 

The program gives two outputs:

- The first output is a TABULAR format file consisting of three columns: 

 - the left column represents the p-values of multiple simultaneous tests, one line for every p-value
 - the middle column represents the q-values corresponding to the p-values
 - the third column represent the significance values, either 1 for significant or 0 for non-significant 

- The second output is a PDF format file consisting of three pages: 

 - the first page represents the p-values histogram
 - the second page represents the q-values histogram
 - the third page represents the four Q-plots as introduced in the "qvalue" package manual.
  

**Example**

Let us have the first input file of p-values as follows::

        0.140627492
        0.432249886
        0.122120877
        0.142010182
        0.012909858
        0.000142807
        0.039841941
        0.035173303
        0.011340057
        1.01E-05
        0.212738282
        0.091256284
        0.547375415
        0.189589833
        6.18E-12
        0.001235875
        1.10E-05
        9.75E-07
        2.13E-18
        2.54E-16
        1.20E-19
        9.76E-14
        0.359181534
        0.03661672
        0.400459987
        0.387436466
        0.342075061
        0.904129283
        0.031152635

Running the program will give the following output::

        pi0: 0.140311054        

        FDR level: 0.05

        p-value         q-value         significant
        0.1406275       0.02889212      1
        0.4322499       0.06514199      0
        0.1221209       0.02760624      1
        0.1420102       0.02889212      1
        0.01290986      0.00437754      1
        0.000142807     6.46E-05        1
        0.03984194      0.01013235      1
        0.0351733       0.009932946     1
        0.01134006      0.004194811     1
        1.01E-05        5.59E-06        1
        0.2127383       0.03934711      1
        0.09125628      0.02184257      1
        0.5473754       0.07954578      0
        0.1895898       0.03673547      1
        6.18E-12        5.03E-12        1
        0.001235875     0.00050288      1
        1.10E-05        5.59E-06        1
        9.75E-07        6.61E-07        1
        2.13E-18        4.33E-18        1
        2.54E-16        3.45E-16        1
        1.20E-19        4.88E-19        1
        9.76E-14        9.93E-14        1
        0.3591815       0.06089654      0
        0.03661672      0.009932946     1
        0.40046 0.0626723       0
        0.3874365       0.0626723       0
        0.3420751       0.06051785      0
        0.9041293       0.1268593       0
        0.03115264      0.009750824     1
        
        
.. image:: ../static/operation_icons/p_hist.png


.. image:: ../static/operation_icons/q_hist.png


.. image:: ../static/operation_icons/Q_plots.png


  </help>  
  
</tool>