root/galaxy-central/tools/ncbi_blast_plus/ncbi_blastp_wrapper.xml @ 3

<tool id="ncbi_blastp_wrapper" name="NCBI BLAST+ blastp" version="0.0.1">
    <description>Search protein database with protein query sequence(s)</description>
    <command>
## The command is a Cheetah template which allows some Python based syntax.
## Lines starting hash hash are comments. Galaxy will turn newlines into spaces
blastp
-query "$query"
#if $db_opts.db_opts_selector == "db":
  -db "$db_opts.database"
#else:
  -subject "$db_opts.subject"
#end if
-task $blast_type
-evalue $evalue_cutoff
$adv_opts.filter_query
$adv_opts.matrix
-out $output1
$out_format
-num_threads 8
## Need int(str(...)) because $adv_opts.max_hits is an InputValueWrapper object not a string
## Note -max_target_seqs overrides -num_descriptions and -num_alignments
#if (str($adv_opts.max_hits) and int(str($adv_opts.max_hits)) > 0):
-max_target_seqs $adv_opts.max_hits
#end if
#if (str($adv_opts.word_size) and int(str($adv_opts.word_size)) > 0):
-word_size $adv_opts.word_size
#end if
##Ungapped disabled for now - see comments below
##$adv_opts.ungapped
    </command>
    <inputs>
        <param name="query" type="data" format="fasta" label="Protein query sequence(s)"/> 
        <conditional name="db_opts">
            <param name="db_opts_selector" type="select" label="Subject database/sequences">
              <option value="db" selected="True">BLAST Database</option>
              <option value="file">FASTA file</option>
            </param>
            <when value="db">
                <param name="database" type="select" label="Protein BLAST database">
                    <options from_file="blastdb_p.loc">
                      <column name="name" index="0"/>
                      <column name="value" index="1"/>
                    </options>
                </param>
                <param name="subject" type="hidden" value="" /> 
            </when>
            <when value="file">
                <param name="database" type="hidden" value="" /> 
                <param name="subject" type="data" format="fasta" label="Protein FASTA file to use as database"/> 
            </when>
        </conditional>
        <param name="blast_type" type="select" display="radio" label="Type of BLAST">
            <option value="blastp">blastp</option>
            <option value="blastp-short">blastp-short</option>
        </param>
        <param name="evalue_cutoff" type="float" size="15" value="0.001" label="set expectation value cutoff" />
        <param name="out_format" type="select" label="Output format">
            <option value="-outfmt 6" selected="True">Tabular</option>
            <option value="-outfmt 5">BLAST XML</option>
            <option value="-outfmt 0">Pairwise text</option>
            <option value="-outfmt 0 -html">Pairwise HTML</option>
            <option value="-outfmt 2">Query-anchored text</option>
            <option value="-outfmt 2 -html">Query-anchored HTML</option>
            <option value="-outfmt 4">Flat query-anchored text</option>
            <option value="-outfmt 4 -html">Flat query-anchored HTML</option>
            <!--
            <option value="-outfmt 11">BLAST archive format (ASN.1)</option>
            -->
        </param>
        <conditional name="adv_opts">
            <param name="adv_opts_selector" type="select" label="Advanced Options">
              <option value="basic" selected="True">Hide Advanced Options</option>
              <option value="advanced">Show Advanced Options</option>
            </param>
            <when value="basic">
                <param name="filter_query" type="hidden" value="" />
                <param name="matrix" type="hidden" value="" />
                <param name="max_hits" type="hidden" value="" />
                <param name="word_size" type="hidden" value="" />
                <!--
                Ungapped disabled for now, see comments below
                <param name="ungapped" type="hidden" value="" />
                -->
            </when>
            <when value="advanced">
                <!-- Could use a select (yes, no, other) where other allows setting 'window locut hicut' -->
                <param name="filter_query" type="boolean" label="Filter out low complexity regions (with SEG)" truevalue="-seg yes" falsevalue="-seg no" checked="true" />
                <param name="matrix" type="select" label="Scoring matrix">
                    <option value="-matrix BLOSUM90">BLOSUM90</option>
                    <option value="-matrix BLOSUM80">BLOSUM80</option>
                    <option value="-matrix BLOSUM62" selected="true">BLOSUM62 (default)</option>
                    <option value="-matrix BLOSUM50">BLOSUM50</option> 
                    <option value="-matrix BLOSUM45">BLOSUM45</option>
                    <option value="-matrix PAM250">PAM250</option>
                    <option value="-matrix PAM70">PAM70</option>
                    <option value="-matrix PAM30">PAM30</option>
                </param>
                <!-- Why doesn't optional override a validator? I want to accept an empty string OR a non-negative integer -->
                <param name="max_hits" type="integer" value="0" label="Maximum hits to show" help="Use zero for default limits">
                    <validator type="in_range" min="0" />
                </param>
                <!-- I'd like word_size to be optional, with minimum 2 for blastp -->
                <param name="word_size" type="integer" value="0" label="Word size for wordfinder algorithm" help="Use zero for default, otherwise minimum 2.">
                    <validator type="in_range" min="0" />
                </param>
                <!--
                Can't use '-ungapped' on its own, error back is:
                Composition-adjusted searched are not supported with an ungapped search, please add -comp_based_stats F or do a gapped search
                Tried using '-ungapped -comp_based_stats F' and blastp crashed with 'Attempt to access NULL pointer.'
                <param name="ungapped" type="boolean" label="Perform ungapped alignment only?" truevalue="-ungapped -comp_based_stats F" falsevalue="" checked="false" />
                -->
            </when>
        </conditional>
    </inputs>
    <outputs>
        <data name="output1" format="tabular" label="${blast_type.value_label} on ${db_opts.db_opts_selector}">
            <change_format>
                <when input="out_format" value="-outfmt 0" format="txt"/>
                <when input="out_format" value="-outfmt 0 -html" format="html"/>
                <when input="out_format" value="-outfmt 2" format="txt"/>
                <when input="out_format" value="-outfmt 2 -html" format="html"/>
                <when input="out_format" value="-outfmt 4" format="txt"/>
                <when input="out_format" value="-outfmt 4 -html" format="html"/>
                <when input="out_format" value="-outfmt 5" format="blastxml"/>
            </change_format>
        </data>
    </outputs>
    <requirements>
        <requirement type="binary">blastn</requirement>
    </requirements>
    <tests>
    </tests>
    <help>
    
.. class:: warningmark

**Note**. Database searches may take substantial amount of time.
For large input datasets it is advisable to allow overnight processing.  

-----

**What it does**

Search a *protein database* using a *protein query*,
using the NCBI BLAST+ blastp command line tool.

-----

**Output format**

Because Galaxy focuses on processing tabular data, the default output of this tool is tabular. This contains 12 columns:

1. Id of your sequence 
2. GI of the database hit 
3. % identity
4. Alignment length
5. # mismatches
6. # gaps
7. Start position in your sequence
8. End position in your sequence
9. Start position in database hit
10. End position in database hit
11. E-value
12. Bit score

The second option is BLAST XML output, which is designed to be parsed by another program, and is understood by other Galaxy tools.

You can also choose several plain text or HTML output formats which are designed to be read by a person (not by another program).
The HTML versions use basic webpage formatting and can include links to the hits on the NCBI website.
The pairwise output (the default on the NCBI BLAST website) shows each match as a pairwise alignment with the query.
The two query anchored outputs show a multiple sequence alignment between the query and all the matches,
and differ in how insertions are shown (marked as insertions or with gap characters added to the other sequences).

-------

**References**

Altschul et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. 1997. Nucleic Acids Res. 25:3389-3402.

Schaffer et al. Improving the accuracy of PSI-BLAST protein database searches with composition-based statistics and other refinements. 2001. Nucleic Acids Res. 29:2994-3005.

    </help>
</tool>