<tool id="rgManQQ1" name="Manhattan/QQ:" version="1.0.3">
    <code file="rgManQQ_code.py"/>

    <description>Plots for WGA P values</description>

    <command interpreter="python">
        rgManQQ.py '$i' "$name" '$out_html' '$out_html.files_path' '$chrom_col' '$offset_col' '$pval_col' '$grey'
    </command>

    <inputs>
      <page>
      <param name="i"  type="data" label="Tabular data from your current history"
      format="tabular" refresh_on_change="true"/>
      </page>
      <page>
       <param name='name' type='text' size="132" value='Manhattan and QQ plots' label="Title for this job"/>
       <param name="pval_col" type='select' size="5" label = 'P value (0-1) column in input file'  
        dynamic_options="get_phecols(i,False,'pval')" refresh_on_change="true" multiple="true" 
        help="(Select multiple P value columns for multiple plots holding down the [Ctrl] key as you click)" />
       <param name="chrom_col" type='select' label = 'Chromosome column in input file'
        help='Select "None" if chromosome not available or no Manhattan plot required'
        dynamic_options="get_phecols(i,True,'chr')" />
       <param name="offset_col" type='select' label = 'Base pair offset column in input file'
        help='Select "None" if offset not available or no Manhattan plot required'
        dynamic_options="get_phecols(i,True,'offs')" />
       <param name="grey" type="boolean" checked="false" truevalue="true" falsevalue="false" 
        label="Grey scale for Manhattan plot (default is colour"/> 
       </page>
    </inputs>
    
   <outputs>
       <data format="html" name="out_html" />
   </outputs>
   <options refresh="True"/>

<tests>
 <test>
 <param name='i' value='smallwgaP.xls' ftype='tabular' >
 </param>
 <param name='name' value='rgManQQtest1' />
 <param name='pval_col' value='7' />
 <param name='chrom_col' value='1' />
 <param name='offset_col' value='2' />
 <param name='grey' value='0' />
 <output name='out_html' file='rgtestouts/rgManQQ/rgManQQtest1.html' ftype='html' lines_diff='60'>
   <extra_files type="file" name='Allelep_manhattan.png' value='rgtestouts/rgManQQ/Allelep_manhattan.png' compare="sim_size" 
     delta = "20000"/>
   <extra_files type="file" name='Allelep_qqplot.png' value='rgtestouts/rgManQQ/Allelep_qqplot.png' compare="sim_size"
     delta = "20000" />
   <extra_files type="file" name='rgManQQtest1.R' value='rgtestouts/rgManQQ/rgManQQtest1.R' compare="diff" lines_diff="160"/>
 </output>
 </test>
</tests>
<help>

.. class:: infomark

**Syntax**

- **Tabular Data** is a tab delimited header file with chromosome, offset and p values to be plotted
- **Chromosome Column** is the column in that data containing the chromosome as an integer
- **Offset Column** contains the offset within the chromosome
- **P Value Column** contains the (untransformed) p values at that locus - choose multiple columns if needed

NOTE - plotting millions of p values may take tens of minutes depending on 
how busy the server is - be patient please. 

-----

.. class:: infomark

**Summary**

This tool will create a qq plot and a Manhattan plot for one or more GWA P value columns from a tabular
dataset. For Manhattan plots, the data must include the chromosome (eg use 23,24,25 for x,y,mt...) and
offset. Many analysis files contain the required fields but even without chromosome and offset, a qq plot 
can be created.

-----

.. class:: infomark

**Explanation**

A "Manhattan" plot shows -log10 p values ordered by offset and by chromosome. Regions with interestingly
improbable p values are above the red line which is drawn at the Bonferroni FWER control level (0.05/n 
where n is the number of tests - this is highly conservative for correlated SNPs typical of GWA)

.. image:: ./static/images/Armitagep_manhattan.png

A quantile-quantile (QQ) plot is a good way to see systematic departures from the null expectation of 
uniform p-values from a genomic analysis. If the QQ plot shows departure from the null (ie a uniform 0-1 
distribution), you hope that this will be in the very smallest p-values suggesting that there might be some 
interesting results to look at. A log scale will help emphasise departures from the null at low p values 
more clear

.. image:: ./static/images/Armitagep_qqplot.png

-----

.. class:: infomark

**Attribution**

This is a Galaxy tool written by Ross Lazarus. It relies on 
ggplot2, an R package from hadley wickham and some 
R code for manhattan and qq plots using ggplot2,
borrowed from Stephen Turner found at http://GettingGeneticsDone.blogspot.com/

copyright Ross Lazarus 2010
Licensed under the terms of the LGPL as documented http://www.gnu.org/licenses/lgpl.html
but is about as useful as a chocolate teapot without R and Galaxy which all have a
twisty maze of little licenses, all different.

I'm no lawyer, but it looks like at least LGPL if you create derived works from this code. 
Good luck.

</help>
</tool>