comparison meme_chip.xml @ 0:6095db402811 draft

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/meme_chip commit 9068afc3812495aaf88c9a2f5a224c634d634742

0:6095db402811

<tool id="meme_chip" name="MEME-ChIP" version="4.11.2">
    <description>- motif discovery, enrichment analysis and clustering on large nucleotide datasets</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <code file="get_meme_motif_databases.py" />
    <command detect_errors="exit_code"><![CDATA[
#import os
#set primary_output = $os.path.join($output.files_path, "index.html")
meme-chip '$input'
-noecho
#if $control:
    -neg '$control'
#end if
$sequence_alphabet
-o '$output.files_path'
#if str($options_type_cond.options_type)=='advanced':
    ## FIXME: CentriMo cannot be run,  See the comments in the input section.
    ## #set run_centrimo = $options_type_cond.run_centrimo_cond.run_centrimo
    ## #if str($run_centrimo) == "yes":
    ##     -db $options_type_cond.run_centrimo_cond.meme_motif_databases.fields.path
    ##     #if $options_type_cond.run_centrimo_cond.centrimo_local:
    ##         -centrimo-local
    ##     #end if
    ##     #if $options_type_cond.run_centrimo_cond.centrimo_score:
    ##         -centrimo-score $options_type_cond.run_centrimo_cond.centrimo_score
    ##     #end if
    ##     #if $options_type_cond.run_centrimo_cond.centrimo_maxreg:
    ##         -centrimo-maxreg $options_type_cond.run_centrimo_cond.centrimo_maxreg
    ##     #end if
    ##     #if $options_type_cond.run_centrimo_cond.centrimo_ethresh:
    ##         -centrimo-ethresh $options_type_cond.run_centrimo_cond.centrimo_ethresh
    ##     #end if
    ##     #if $options_type_cond.run_centrimo_cond.centrimo_noseq:
    ##         -centrimo-noseq
    ##     #end if
    ##     #if $options_type_cond.run_centrimo_cond.centrimo_flip:
    ##         -centrimo-flip
    ##     #end if
    ## #end if
    $options_type_cond.search_given_strand
    -order $options_type_cond.background_model_order
    #if str($options_type_cond.subsampling_cond.subsampling) == "no":
        -norand
        #if $options_type_cond.subsampling_cond.subsampling.seed:
            -seed $options_type_cond.subsampling_cond.subsampling.seed
        #end if
    #end if
    #if $options_type_cond.nmeme:
        -nmeme $options_type_cond.nmeme
    #end if
    #if $options_type_cond.ccut:
        -ccut $options_type_cond.ccut
    #end if
    -group-thresh $options_type_cond.group_threash
    #if str($options_type_cond.group_weak):
        -group-weak $options_type_cond.group_weak
    #end if
    -filter-thresh $options_type_cond.filter_thresh
    $options_type_cond.old_clustering
    -meme-mod $options_type_cond.meme_mod
    #if $options_type_cond.meme_minw:
        -meme-minw $options_type_cond.meme_minw
    #end if
    #if $options_type_cond.meme_maxw:
        -meme-maxw $options_type_cond.meme_maxw
    #end if
    #if $options_type_cond.meme_nmotifs:
        -meme-nmotifs $options_type_cond.meme_nmotifs
    #end if
    #if $options_type_cond.meme_minsites:
        -meme-minsites $options_type_cond.meme_minsites
    #end if
    #if $options_type_cond.meme_maxsites:
        -meme-maxsites $options_type_cond.meme_maxsites
    #end if
    $options_type_cond.meme_pal
    -dreme-e $options_type_cond.dreme_e
    -dreme-m $options_type_cond.dreme_m
    -spamo-skip
    -fimo-skip
#end if
&& rm '$output'
&& ln -s $primary_output '$output'
    ]]></command>
    <inputs>
        <param name="input" type="data" format="fasta" label="Primary sequences" help="Nucleotide sequences must have equal length"/>
        <param name="control" type="data" format="fasta" optional="true" label="Control sequences" help="If no selection, positive sequences in the input are shuffled to create the negative set"/>
        <param name="sequence_alphabet" type="select" label="Sequence alphabet">
            <option value="-dna" selected="true">DNA</option>
            <option value="-rna">RNA</option>
        </param>
        <conditional name="options_type_cond">
            <param name="options_type" type="select" label="Options Configuration">
                <option value="basic" selected="true">Basic</option>
                <option value="advanced">Advanced</option>
            </param>
            <when value="basic"/>
            <when value="advanced">
                <!--
                FIXME: CentriMo cannot be run since the tool form cannot populate the mem_motif_database select list below.
                <conditional name="run_centrimo_cond">
                    <param name="run_centrimo" type="select" label="Run TOMTOM and CentriMo?">
                        <option value="yes" selected="true">Yes</option>
                        <option value="no">No</option>
                    </param>
                    <when value="yes">

                        We have 2 dynamic select lists here.  The first select list (meme_motif_database_dir) is populated from the meme_motif_databases
                        data table.  The second select list (meme_motif_database) is dynamically re-rendered whenever the selection in the meme_motif_database_dir
                        select list is changed.  This composition used to work (see Examples->Dynamic Options section of
                        https://docs.galaxyproject.org/en/latest/dev/schema.html) but no longer does.  We'll have to figure out what is broken in
                        the dynamic options code in ~/parameters/basic.py in order to uncomment this block.

                        <param name="meme_motif_database_dir" type="select" label="Select the motifs (DNA)" refresh_on_change="True">
                            <options from_data_table="meme_motif_databases">
                                <filter type="sort_by" column="1"/>
                                <validator type="no_options" message="No MEME motif databases are available for the selected input"/>
                            </options>
                        </param>
                        <param name="meme_motif_database" type="select" label="MEME motif database" dynamic_options="get_meme_motif_database_options(file_path=meme_motif_database_dir)"/>
                        <param name="centrimo_local" type="boolean" truevalue="true" falsevalue="" checked="False" label="Compute enrichment of all regions"/>
                        <param name="centrimo_score" type="integer" optional="true" value="0" min="0" label="Minimum allowed CentriMo match score"/>
                        <param name="centrimo_maxreg" type="integer" optional="true" value="0" min="0" label="Maximum CentriMo region size to be considered"/>
                        <param name="centrimo_ethresh" type="integer" optional="true" value="0" min="0" label="CentriMo E-value threshold for reporting" />
                        <param name="centrimo_noseq" type="boolean" truevalue="true" falsevalue="" checked="False" label="Store CentriMo sequence IDs in the output"/>
                        <param name="centrimo_flip" type="boolean" truevalue="true" falsevalue="" checked="False" label="Reflect CentriMo matches on reverse strand around center"/>
                    </when>
                    <when value="no"/>
                </conditional>
                -->
                <param name="background_model_order" type="select" label="Select the order of the Markov background model">
                    <option value="0">0-order model of sequences</option>
                    <option value="1" selected="True">1st order model of sequences</option>
                    <option value="2">2nd order model of sequences</option>
                    <option value="3">3rd order model of sequences</option>
                    <option value="4">4th order model of sequences</option>
                </param>
                <param name="nmeme" type="integer" optional="true" value="" min="1" label="Limit of sequences to pass to MEME"/>
                <conditional name="subsampling_cond">
                    <param name="subsampling" type="select" label="Should subsampling be random?" help="Select 'No' if your input sequences are sorted in order of confidence (best to worst)">
                        <option value="yes" selected="true">Yes</option>
                        <option value="no">No</option>
                    </param>
                    <when value="yes">
                        <param name="seed" type="integer" optional="true" value="" min="1" label="Seed for the randomized selection of sequences"/>
                    </when>
                    <when value="no"/>
                </conditional>
                <param name="ccut" type="integer" optional="true" value="100" min="0" label="maximum size of a sequence before it is cut down to a centered section" help="Zero value indicates the sequences should not be cut down"/>
                <param name="group_threash" type="float" value="0.05" min="0" label="Primary threshold for clustering motifs" />
                <param name="group_weak" type="float" optional="true" value="0" min="0" label="Secondary threshold for clustering motifs" help="Zero value results in 2*primary threshold"/>
                <param name="filter_thresh" type="float" value="0.05" min="0" label="E-value threshold for including motifs"/>
                <param name="search_given_strand" type="boolean" truevalue="-norc" falsevalue="" checked="False" label="Search given strand only"/>
                <param argument="-old_clustering" type="boolean" truevalue="-old_clustering" falsevalue="" checked="False" label="Pick cluster seed motifs based only on significance"/>
                <param name="meme_mod" type="select" label="What is the expected motif site distribution?">
                    <option value="oops" selected="True">One occurance per sequence</option>
                    <option value="zoops">Zero or one occurances per sequence</option>
                    <option value="anr">Any number of repititions</option>
                </param>
                <param name="meme_minw" type="integer" optional="true" value="0" min="0" label="Minimum motif width"/>
                <param name="meme_maxw" type="integer" optional="true" value="0" min="0" label="Maximum motif width"/>
                <param name="meme_nmotifs" type="integer" optional="true" value="0" min="0" label="Maximum number of motifs to find"/>
                <param name="meme_minsites" type="integer" optional="true" value="0" min="0" label="Minimum number of sites per motif"/>
                <param name="meme_maxsites" type="integer" optional="true" value="0" label="Maximum number of sites per motif"/>
                <param argument="-meme_pal" type="boolean" truevalue="-meme-pal" falsevalue="" checked="False" label="Look for palindromes only"/>
                <param name="dreme_e" type="float" value="0.05" min="0" label="Stop DREME searching after reaching this E-value threshold"/>
                <param name="dreme_m" type="integer" value="10" min="1" label="Stop DREME searching after finding this many motifs" />
            </when>
        </conditional>
        <param name="non_commercial_use" label="I certify that I am not using this tool for commercial purposes." type="boolean" truevalue="NON_COMMERCIAL_USE" falsevalue="COMMERCIAL_USE" checked="False">
            <validator type="expression" message="This tool is only available for non-commercial use.">value == True</validator>
        </param>
    </inputs>
    <outputs>
        <data name="output" format="html" label="${tool.name} (html) on ${on_string}"/>
    </outputs>
    <tests>
        <test>
            <param name="input" value="input1.fasta" ftype="fasta"/>
            <param name="non_commercial_use" value="True"/>
            <output name="output" file="output1.html" ftype="html" compare="contains"/>
        </test>
        <test>
            <param name="input" value="input1.fasta" ftype="fasta"/>
            <param name="sequence_alphabet" value="-rna"/>
            <param name="options_type" value="advanced"/>
            <param name="background_model_order" value="0"/>
            <param name="non_commercial_use" value="True"/>
            <output name="output" file="output1.html" ftype="html" compare="contains"/>
        </test>
    </tests>
    <help>
.. class:: warningmark

**WARNING: This tool is only available for non-commercial use. Use for educational, research and non-profit purposes is permitted.
Before using, be sure to review, agree, and comply with the license.**

MWMW-ChIP perform motif discovery, motif enrichment analysis and clustering on large nucleotide datasets.

If you want to specify sequence weights, you must include them at the top of your input FASTA file.

MEME discovers novel, ungapped motifs (recurring, fixed-length patterns) in your sequences (sample output from sequences).
MEME splits variable-length patterns into two or more separate motifs.  A motif is a sequence pattern that occurs repeatedly
in a group of related sequences.  MEME represents motifs as position-dependent letter-probability matrices which describe the
probability of each possible letter at each position in the pattern.  Individual MEME motifs do not contain gaps.  Patterns
with variable-length gaps are split by MEME into two or more separate motifs.  MEME takes as input a group of sequences and
outputs as many motifs as requested.  MEME uses statistical modeling techniques to automatically choose the best width, number
of occurrences, and description for each motif.

.. class:: infomark

For detailed information on MEME, click here_, or view the license_.

.. _here: http://meme-suite.org/doc/meme.html?man_type=web
.. _license: http://meme-suite.org/doc/copyright.html?man_type=web

    </help>
    <citations>
        <citation type="doi">10.1093/bioinformatics/btr189</citation>
    </citations>
</tool>