<tool id="gd_calc_freq" name="Rank Pathways" version="1.2.0">
  <description>: Assess the impact of a gene set on KEGG pathways</description>

  <command interpreter="python">
    #if $rank_by.choice == 'pct'
      rank_pathways_pct.py
      --input '$rank_by.input1'
      --columnENSEMBLT '$rank_by.t_col1'
      --inBckgrndfile '$rank_by.input2'
      --columnENSEMBLTBckgrnd '$rank_by.t_col2'
      --columnKEGGBckgrnd '$rank_by.k_col2'
      --statsTest '$rank_by.stat'
      --output '$output'
    #else if $rank_by.choice == 'paths'
      calclenchange.py
      '--loc_file=${GALAXY_DATA_INDEX_DIR}/gd.rank.loc'
      '--species=${rank_by.input.metadata.dbkey}'
      '--input=${rank_by.input}'
      '--output=${output}'
      '--posKEGGclmn=${rank_by.kpath}'
      '--KEGGgeneposcolmn=${rank_by.kgene}'
    #end if
  </command>

  <inputs>
    <conditional name="rank_by">
      <param name="choice" type="select" label="Rank by">
        <option value="pct" selected="true">percentage of genes affected</option>
        <option value="paths">change in length and number of paths</option>
      </param>
      <when value="pct">
        <!-- using fields similar to the Rank Terms tool -->
        <param name="input1" type="data" format="tabular" label="Query dataset" />
        <param name="t_col1" type="data_column" data_ref="input1" label="Column with ENSEMBL transcript codes" />
        <param name="input2" type="data" format="tabular" label="Background dataset" />
        <param name="t_col2" type="data_column" data_ref="input2" label="Column with ENSEMBL transcript codes" />
        <param name="k_col2" type="data_column" data_ref="input2" label="Column with KEGG pathways" />
        <param name="stat" type="select" label="Statistic for determining enrichment/depletion">
          <option value="fisher" selected="true">two-tailed Fisher's exact test</option>
          <option value="hypergeometric">hypergeometric test</option>
          <option value="binomial">binomial probability</option>
        </param>
      </when>
      <when value="paths">
        <param name="input" type="data" format="tabular" label="Dataset" />
        <param name="kgene" type="data_column" data_ref="input" label="Column with KEGG gene ID" />
        <param name="kpath" type="data_column" data_ref="input" numerical="false" label="Column with KEGG pathways" />
      </when>
    </conditional>
  </inputs>

  <outputs>
    <data name="output" format="tabular" />
  </outputs>

  <tests>
    <test>
    </test>
  </tests>

  <help>

**Dataset formats**

All of the input and output datasets are in tabular_ format.
The input dataset must have columns with KEGG gene ID and pathways.
[Need to update this, since input columns now depend on the "Rank by" choice.]
The output datasets are described below.
(`Dataset missing?`_)

.. _tabular: ./static/formatHelp.html#tab
.. _Dataset missing?: ./static/formatHelp.html

-----

**What it does**

This tool produces a table ranking the pathways based on the percentage
of genes in an input dataset, out of the total in each pathway
[please clarify w.r.t. query and background datasets].
Alternatively, the tool ranks the pathways based on the change in
length and number of paths connecting sources and sinks.  This change is
calculated between graphs representing pathways with and without excluding
the nodes that represent the genes in an input list.  Sources are all
the nodes representing the initial reactants/products in the pathway.
Sinks are all the nodes representing the final reactants/products in
the pathway.

If pathways are ranked by percentage of genes affected, the output contains
a row for each KEGG pathway, with the following columns:

1. count: the number of genes in the query set that are in this pathway
2. representation: the percentage of this pathway's genes (from the background dataset) that appear in the query set
3. ranking of this pathway, based on its representation ("1" is highest)
4. probability of depletion of this pathway in the query dataset
5. probability of enrichment of this pathway in the query dataset
6. KEGG pathway

If pathways are ranked by change in length and number of paths, the
output is a tabular dataset with the following columns:

1. change in the mean length of paths between sources and sinks
2. mean length of paths between sources and sinks in the pathway including the genes in the input dataset.  If the pathway do not have sources/sinks, the length is assumed to be infinite (I)
3. mean length of paths between sources and sinks in the pathway excluding the genes in the input dataset.  If the pathway do not have sources/sinks, the length is assumed to be infinite (I)
4. rank of the change in the mean length of paths between sources and sinks (from high change to low change)
5. change in the number of paths between sources and sinks
6. number of paths between sources and sinks in the pathway including the genes in the input dataset.  If the pathway do not have sources/sinks, it is assumed to be a circuit (C)
7. number of paths between sources and sinks in the pathway excluding the genes in the input dataset.  If the pathway do not have sources/sinks, it is assumed to be a circuit (C)
8. rank of the change in the number of paths between sources and sinks (from high change to low change)
9. name of the pathway

-----

**Examples**

- input (column 10 for KEGG gene ID, column 12 for KEGG pathways)::
 
   Contig39_chr1_3261104_3261850   414  chr1  3261546  ENSCAFT00000000001   ENSCAFP00000000001   S    667   F    476153  probably damaging    cfa00230=Purine metabolism.cfa00500=Starch and sucrose metabolism.cfa00740=Riboflavin metabolism.cfa00760=Nicotinate and nicotinamide metabolism.cfa00770=Pantothenate and CoA biosynthesis.cfa01100=Metabolic pathways
   Contig62_chr1_19011969_19012646 265  chr1  19012240 ENSCAFT00000000144   ENSCAFP00000000125   *    161   R    483960  probably damaging    N
   etc.
 
- output ranked by percentage of genes affected [need new sample output with more columns]::

   3   0.25   1   cfa03450=Non-homologous end-joining
   1   0.25   1   cfa00750=Vitamin B6 metabolism
   2   0.2    3   cfa00290=Valine, leucine and isoleucine biosynthesis
   3   0.18   4   cfa00770=Pantothenate and CoA biosynthesis
   etc.

- output ranked by change in length and number of paths::

    3.64   8.44   4.8     2   4    9    5   1   cfa00260=Glycine, serine and threonine metabolism
    7.6    9.6    2       1   3    5    2   2   cfa00240=Pyrimidine metabolism
    0.05   2.67   2.62    6   1   30   29   3   cfa00982=Drug metabolism - cytochrome P450
   -0.08   8.33   8.41   84   1   30   29   3   cfa00564=Glycerophospholipid metabolism
   etc.

  </help>
</tool>