<tool id="Psortb" name="psortb" version="0.0.1">
  <description>Determines sub-cellular localisation of bacterial/archaeal protein sequences</description>
  <!-- If job splitting is enabled, break up the query file into parts -->
  <!-- Using 2000 chunks meaning 4 threads doing 500 each is ideal -->
  <parallelism method="basic" split_inputs="fasta_file" split_mode="to_size" split_size="2000" merge_outputs="tabular_file"></parallelism>
  <version_command interpreter="python">psortb.py --version</version_command>
  <command interpreter="python">psortb.py "\$NSLOTS" "$type" "$long" "$cutoff" "$divergent" "$sequence" "$outfile"</command>
  <stdio>
    <!-- Anything other than zero is an error -->
    <exit_code range="1:" />
    <exit_code range=":-1" />
  </stdio>
  <inputs>
    <param format="fasta" name="sequence" type="data"
	   label="Input sequences for which to predict localisation (protein FASTA format)" />
    <param name="type" type="select"
	   label="Organism type (N.B. all sequences in the above file must be of the same type)" >
      <option value="-p">Gram positive bacteria</option>
      <option value="-n">Gram negative bacteria</option>
      <option value="-a">Archaea</option>
    </param>
    <param name="long" type="select" label="Output type">
      <option value="terse">Short (terse, tabular with 3 columns)</option>
      <!-- The normal output is text, not tabular - worth offering?
      <option value="normal">Normal</option>
      -->
      <option value="long">Long (verbose, tabular with about 30 columns, depending on organism type)</option>
    </param>
    <param name="cutoff" size="10" type="float" optional="true" value=""
	   label="Sets a cutoff value for reported results (e.g. 7.5)"
	   help="Leave blank or use zero for no cutoff." />
    <param name="divergent" size="10" type="float" optional="true" value=""
	   label="Sets a cutoff value for the multiple localization flag (e.g. 4.5)"
	   help="Leave blank or use zero for no cutoff." />
  </inputs>
  <outputs>
    <data format="tabular" name="outfile" />
  </outputs>
  <requirements>
    <requirement type="binary">psort</requirement>
  </requirements>
  <tests>
    <test>
      <param name="sequence" value="empty.fasta" ftype="fasta"/>
      <param name="long" value="terse"/>
      <output name="outfile" file="empty_psortb_terse.tabular" ftype="tabular"/>
    </test>
    <test>
      <param name="sequence" value="k12_ten_proteins.fasta" ftype="fasta"/>
      <param name="long" value="terse"/>
      <output name="outfile" file="k12_ten_proteins_psortb_p_terse.tabular" ftype="tabular"/>
    </test>
  </tests>
  <help>

**What it does**

This calls the command line tool PSORTb v3.0 for prediction of prokaryotic
localization sites. The input dataset needs to be protein FASTA sequences.
The default output is a simple tabular file with three columns, one row
per query sequence:

====== ==============================
Column Description
------ ------------------------------
     1 Sequence identifier
     2 Localisation, e.g. Cytoplasmic
     3 Score
====== ==============================

The long output is also tabular with one row per query sequence, but has
lots more columns (a different set for each supported organism type). In
both cases, a simple header line is included (starting with a hash, #,
so that Galaxy treats it as a comment) giving the column names.


**References**

N.Y. Yu, J.R. Wagner, M.R. Laird, G. Melli, S. Rey, R. Lo, P. Dao,
S.C. Sahinalp, M. Ester, L.J. Foster, F.S.L. Brinkman (2010)
PSORTb 3.0: Improved protein subcellular localization prediction with
refined localization subcategories and predictive capabilities for all
prokaryotes, Bioinformatics 26(13):1608-1615
http://dx.doi.org/10.1093/bioinformatics/btq249

http://www.psort.org/documentation/index.html

  </help>
</tool>