annotate tools/ncbi_blast_plus/ncbi_tblastn_wrapper.xml @ 11:4c4a0da938ff draft

Uploaded v0.0.22, now wraps BLAST+ 2.2.28 allowing extended tabular output to include the hit descriptions as column 25. Supports $GALAXY_SLOTS. Includes more tests and heavy use of macros.

<tool id="ncbi_tblastn_wrapper" name="NCBI BLAST+ tblastn" version="0.0.22">
    <description>Search translated nucleotide database with protein query sequence(s)</description>
    <!-- If job splitting is enabled, break up the query file into parts -->
    <parallelism method="multi" split_inputs="query" split_mode="to_size" split_size="1000" merge_outputs="output1"></parallelism>
    <macros>
        <token name="@BINARY@">tblastn</token>
        <import>ncbi_macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <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
tblastn
-query "$query"
@BLAST_DB_SUBJECT@
-evalue $evalue_cutoff
@BLAST_OUTPUT@
@THREADS@
#if $adv_opts.adv_opts_selector=="advanced":
-db_gencode $adv_opts.db_gencode
-matrix $adv_opts.matrix
@ADVANCED_OPTIONS@
##Ungapped disabled for now - see comments below
##$adv_opts.ungapped
## End of advanced options:
#end if
    </command>

    <expand macro="stdio" />

    <inputs>
        <param name="query" type="data" format="fasta" label="Protein query sequence(s)"/> 

        <expand macro="input_conditional_nucleotide_db" />
        <expand macro="input_out_format" />
        <expand macro="input_evalue" />
        <expand macro="advanced_options">
            <expand macro="input_db_gencode" />

            <!-- Could use a select (yes, no, other) where other allows setting 'window locut hicut' -->
            <expand macro="input_filter_query_default_true" />
            <expand macro="input_scoring_matrix" />
            <expand macro="input_max_hits" />
            <expand macro="input_word_size" />
            <!--
            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 tblastn 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" />
            -->
            <expand macro="input_parse_deflines" />
        </expand>
    </inputs>
    <outputs>
        <data name="output1" format="tabular" label="tblastn $query.name vs @ON_DB_SUBJECT@">
            <expand macro="output_change_format" />
        </data>
    </outputs>
    <tests>
        <test>
            <param name="query" value="four_human_proteins.fasta" ftype="fasta" />
            <param name="db_opts_selector" value="file" />
            <param name="subject" value="rhodopsin_nucs.fasta" ftype="fasta" />
            <param name="database" value="" />
            <param name="evalue_cutoff" value="1e-10" />
            <param name="out_format" value="5" />
            <param name="adv_opts_selector" value="advanced" />
            <param name="filter_query" value="false" />
            <param name="matrix" value="BLOSUM80" />
            <param name="max_hits" value="0" />
            <param name="word_size" value="0" />
            <param name="parse_deflines" value="false" />
            <output name="output1" file="tblastn_four_human_vs_rhodopsin.xml" ftype="blastxml" />
        </test>
        <test>
            <param name="query" value="four_human_proteins.fasta" ftype="fasta" />
            <param name="db_opts_selector" value="file" />
            <param name="subject" value="rhodopsin_nucs.fasta" ftype="fasta" />
            <param name="database" value="" />
            <param name="evalue_cutoff" value="1e-10" />
            <param name="out_format" value="ext" />
            <param name="adv_opts_selector" value="advanced" />
            <param name="filter_query" value="false" />
            <param name="matrix" value="BLOSUM80" />
            <param name="max_hits" value="0" />
            <param name="word_size" value="0" />
            <param name="parse_deflines" value="false" />
            <output name="output1" file="tblastn_four_human_vs_rhodopsin_ext.tabular" ftype="tabular" />
        </test>
        <test>
            <param name="query" value="four_human_proteins.fasta" ftype="fasta" />
            <param name="db_opts_selector" value="file" />
            <param name="subject" value="rhodopsin_nucs.fasta" ftype="fasta" />
            <param name="database" value="" />
            <param name="evalue_cutoff" value="1e-10" />
            <param name="out_format" value="6" />
            <param name="adv_opts_selector" value="advanced" />
            <param name="filter_query" value="false" />
            <param name="matrix" value="BLOSUM80" />
            <param name="max_hits" value="0" />
            <param name="word_size" value="0" />
            <param name="parse_deflines" value="false" />
            <output name="output1" file="tblastn_four_human_vs_rhodopsin.tabular" ftype="tabular" />
        </test>
        <test>
            <!-- Same as above, but parse deflines - on BLAST 2.2.25+ - 2.2.27+ makes no difference -->
            <param name="query" value="four_human_proteins.fasta" ftype="fasta" />
            <param name="db_opts_selector" value="file" />
            <param name="subject" value="rhodopsin_nucs.fasta" ftype="fasta" />
            <param name="database" value="" />
            <param name="evalue_cutoff" value="1e-10" />
            <param name="out_format" value="6" />
            <param name="adv_opts_selector" value="advanced" />
            <param name="filter_query" value="false" />
            <param name="matrix" value="BLOSUM80" />
            <param name="max_hits" value="0" />
            <param name="word_size" value="0" />
            <param name="parse_deflines" value="true" />
            <output name="output1" file="tblastn_four_human_vs_rhodopsin.tabular" ftype="tabular" />
        </test>
        <test>
            <param name="query" value="four_human_proteins.fasta" ftype="fasta" />
            <param name="db_opts_selector" value="file" />
            <param name="subject" value="rhodopsin_nucs.fasta" ftype="fasta" />
            <param name="database" value="" />
            <param name="evalue_cutoff" value="1e-10" />
            <param name="out_format" value="0 -html" />
            <param name="adv_opts_selector" value="advanced" />
            <param name="filter_query" value="false" />
            <param name="matrix" value="BLOSUM80" />
            <param name="max_hits" value="0" />
            <param name="word_size" value="0" />
            <param name="parse_deflines" value="false" />
            <output name="output1" file="tblastn_four_human_vs_rhodopsin.html" ftype="html" />
        </test>
    </tests>
    <help>
    
@SEARCH_TIME_WARNING@

**What it does**

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

@FASTA_WARNING@

-----

@OUTPUT_FORMAT@

-------

**References**

If you use this Galaxy tool in work leading to a scientific publication please
cite the following papers:

@REFERENCES@
    </help>
</tool>