annotate tree.xml @ 2:ce3fd529a6ec draft default tip

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/checkm commit 3ae2bd72f789518e95ef65b97e0e5ac90165e113

<tool id="checkm_tree" name="CheckM tree" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="@PROFILE@">
    <description>
        Place bins in the genome tree
    </description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="biotools"/>
    <expand macro="requirements"/>
    <expand macro="version"/>
    <command detect_errors="exit_code"><![CDATA[
@BIN_INPUTS@

checkm tree
    'bins'
    'output'
    $reduced_tree
    $ali
    $nt
    $genes
    --extension 'fasta'
    --threads \${GALAXY_SLOTS:-1}
    --pplacer_threads \${GALAXY_SLOTS:-1}
    ]]></command>
    <inputs>
        <expand macro="bin_inputs" />
        <expand macro="tree_params" />
        <param name="extra_outputs" type="select" multiple="true" optional="true" label="Extra outputs">
            <expand macro="tree_extra_output_options" />
        </param>
    </inputs>
    <outputs>
        <data name="phylo_hmm_info" format="zip" from_work_dir="output/storage/phylo_hmm_info.pkl.gz" label="${tool.name} on ${on_string}: Phylogenetic HMM model info for each bin" />
        <data name="bin_stats_tree" format="tabular" from_work_dir="output/storage/bin_stats.tree.tsv" label="${tool.name} on ${on_string}: Phylogenetic bin stats" />
        <collection name="hmmer_tree" type="list" label="${tool.name} on ${on_string}: Phylogenetic HMM hits to each bin">
            <discover_datasets pattern="(?P&lt;designation&gt;.*)/hmmer\.tree\.txt" format="txt" directory="output/bins/" recurse="true" match_relative_path="true"/>
        </collection>
        <data name="concatenated_fasta" format="fasta" from_work_dir="output/storage/tree/concatenated.fasta" label="${tool.name} on ${on_string}: Concatenated masked sequences"/>
        <data name="concatenated_tre" format="phyloxml" from_work_dir="output/storage/tree/concatenated.tre" label="${tool.name} on ${on_string}: Concatenated tree"/>
        <collection name="hmmer_tree_ali" type="list" label="${tool.name} on ${on_string}: Phylogenetic HMMER alignment file for each bin">
            <filter>ali and extra_outputs and 'hmmer_tree_ali' in extra_outputs</filter>
            <discover_datasets pattern="(?P&lt;designation&gt;.*)/hmmer\.tree\.ali\.txt" format="txt" directory="output/bins/" recurse="true" match_relative_path="true"/>
        </collection>
        <data name="concatenated_pplacer_json" format="json" from_work_dir="output/storage/tree/concatenated.pplacer.json" label="${tool.name} on ${on_string}: Concatenated pplacer JSON">
            <filter>extra_outputs and 'concatenate_pplacer_json' in extra_outputs</filter>
        </data>
        <collection name="genes_fna" type="list" label="${tool.name} on ${on_string}: Protein gene sequences for each bin">
            <filter>not genes and nt and extra_outputs and 'genes_fna' in extra_outputs</filter>
            <discover_datasets pattern="(?P&lt;designation&gt;.*)/genes\.fna" format="fasta" directory="output/bins/" recurse="true" match_relative_path="true"/>
        </collection>
        <collection name="genes_faa" type="list" label="${tool.name} on ${on_string}: Nucleotide gene sequences for each bin">
            <filter>extra_outputs and 'genes_faa' in extra_outputs</filter>
            <discover_datasets pattern="(?P&lt;designation&gt;.*)/genes\.faa" format="fasta" directory="output/bins/" recurse="true" match_relative_path="true"/>
        </collection>
        <collection name="genes_gff" type="list" label="${tool.name} on ${on_string}: Gene feature files for each bin">
            <filter>not genes and extra_outputs and 'genes_gff' in extra_outputs</filter>
            <discover_datasets pattern="(?P&lt;designation&gt;.*)/genes\.gff" format="gff" directory="output/bins/" recurse="true" match_relative_path="true"/>
        </collection>
    </outputs>
    <tests>
        <test expect_num_outputs="6">
            <conditional name="bins">
                <param name="select" value="collection"/>
                <param name="bins_coll">
                    <collection type="list">
                        <element name="637000110" ftype="fasta" value="637000110.fna"/>
                    </collection>
                </param>
            </conditional>
            <param name="reduced_tree" value="true"/>
            <param name="ali" value="false"/>
            <param name="nt" value="false"/>
            <param name="genes" value="false"/>       
            <param name="extra_outputs" value="genes_gff"/>
            <output name="phylo_hmm_info" ftype="zip">
                <assert_contents>
                    <has_size value="1579" delta="10"/>
                </assert_contents>
            </output>
            <output name="bin_stats_tree" ftype="tabular">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="Coding density"/>
                    <has_text text="# predicted genes"/>
                </assert_contents>
            </output>
            <output_collection name="hmmer_tree" count="1">
                <element name="637000110" ftype="txt">
                    <assert_contents>
                        <has_text text="target name"/>
                        <has_text text="AC_000091_551"/>
                    </assert_contents>
                </element>
            </output_collection>
            <output name="concatenated_fasta" ftype="fasta">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="MLKAGVHFGHQTRYW"/>
                </assert_contents>
            </output>
            <output name="concatenated_tre" ftype="phyloxml">
                <assert_contents>
                    <has_text text="IMG_646564547"/>
                    <has_text text="g__Methanocaldococcus"/>
                </assert_contents>
            </output>
            <output_collection name="genes_gff" count="1">
                <element name="637000110" ftype="gff">
                    <assert_contents>
                        <has_text text="#gff-version"/>
                        <has_text text="AC_000091"/>
                        <has_text text="Prodigal.v2.6.3"/>
                    </assert_contents>
                </element>
            </output_collection>
        </test>
        <test expect_num_outputs="8">
            <conditional name="bins">
                <param name="select" value="collection"/>
                <param name="bins_coll">
                    <collection type="list">
                        <element name="637000110" ftype="fasta" value="637000110.fna"/>
                    </collection>
                </param>
            </conditional>
            <param name="reduced_tree" value="true"/>
            <param name="ali" value="true"/>
            <param name="nt" value="false"/>
            <param name="genes" value="false"/>
            <param name="extra_outputs" value="hmmer_tree_ali,concatenate_pplacer_json,genes_faa"/>
            <output name="phylo_hmm_info" ftype="zip">
                <assert_contents>
                    <has_size value="1579" delta="10"/>
                </assert_contents>
            </output>
            <output name="bin_stats_tree" ftype="tabular">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="Coding density"/>
                    <has_text text="# predicted genes"/>
                </assert_contents>
            </output>
            <output_collection name="hmmer_tree" count="1">
                <element name="637000110" ftype="txt">
                    <assert_contents>
                        <has_text text="target name"/>
                        <has_text text="AC_000091_551"/>
                    </assert_contents>
                </element>
            </output_collection>
            <output name="concatenated_fasta" ftype="fasta">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="MLKAGVHFGHQTRYW"/>
                </assert_contents>
            </output>
            <output name="concatenated_tre" ftype="phyloxml">
                <assert_contents>
                    <has_text text="IMG_646564547"/>
                    <has_text text="g__Methanocaldococcus"/>
                </assert_contents>
            </output>
            <output_collection name="hmmer_tree_ali" count="1">
                <element name="637000110" ftype="txt">
                    <assert_contents>
                        <has_text text="hmmsearch"/>
                        <has_text text="Query"/>
                        <has_text text="Ribosom_S12_S23"/>
                    </assert_contents>
                </element>
            </output_collection>
            <output name="concatenated_pplacer_json" ftype="json">
                <assert_contents>
                    <has_text text="IMG_646564547"/>
                    <has_text text="f__Methanocaldococcaceae"/>
                    <has_text text="tree"/>
                </assert_contents>
            </output>
            <output_collection name="genes_faa" count="1">
                <element name="637000110" ftype="fasta">
                    <assert_contents>
                        <has_text text=">AC_000091_1"/>
                        <has_text text="AFLDYFAD"/>
                        <has_text text=">AC_000091_2"/>
                    </assert_contents>
                </element>
            </output_collection>
        </test>
        <test expect_num_outputs="6">
            <conditional name="bins">
                <param name="select" value="collection"/>
                <param name="bins_coll">
                    <collection type="list">
                        <element name="637000110" ftype="fasta" value="637000110.fna"/>
                    </collection>
                </param>
            </conditional>
            <param name="reduced_tree" value="true"/>
            <param name="ali" value="false"/>
            <param name="nt" value="true"/>
            <param name="genes" value="false"/>       
            <param name="extra_outputs" value="genes_fna"/>
            <output name="phylo_hmm_info" ftype="zip">
                <assert_contents>
                    <has_size value="1579" delta="10"/>
                </assert_contents>
            </output>
            <output name="bin_stats_tree" ftype="tabular">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="Coding density"/>
                    <has_text text="# predicted genes"/>
                </assert_contents>
            </output>
            <output_collection name="hmmer_tree" count="1">
                <element name="637000110" ftype="txt">
                    <assert_contents>
                        <has_text text="target name"/>
                        <has_text text="AC_000091_551"/>
                    </assert_contents>
                </element>
            </output_collection>
            <output name="concatenated_fasta" ftype="fasta">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="MLKAGVHFGHQTRYW"/>
                </assert_contents>
            </output>
            <output name="concatenated_tre" ftype="phyloxml">
                <assert_contents>
                    <has_text text="IMG_646564547"/>
                    <has_text text="g__Methanocaldococcus"/>
                </assert_contents>
            </output>
            <output_collection name="genes_fna" count="1">
                <element name="637000110" ftype="fasta">
                    <assert_contents>
                        <has_text text=">AC_000091_1"/>
                        <has_text text="GCGTTTCTGGATT"/>
                    </assert_contents>
                </element>
            </output_collection>
        </test>
        <test expect_num_outputs="5">
            <conditional name="bins">
                <param name="select" value="collection"/>
                <param name="bins_coll">
                    <collection type="list">
                        <element name="637000110" ftype="fasta" value="637000110.faa"/>
                    </collection>
                </param>
            </conditional>
            <param name="reduced_tree" value="true"/>
            <param name="ali" value="false"/>
            <param name="nt" value="false"/>
            <param name="genes" value="true"/>
            <param name="extra_outputs" value=""/>
            <output name="phylo_hmm_info" ftype="zip">
                <assert_contents>
                    <has_size value="1579" delta="10"/>
                </assert_contents>
            </output>
            <output name="bin_stats_tree" ftype="tabular">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="Coding density"/>
                    <has_text text="# predicted genes"/>
                </assert_contents>
            </output>
            <output_collection name="hmmer_tree" count="1">
                <element name="637000110" ftype="txt">
                    <assert_contents>
                        <has_text text="target name"/>
                        <has_text text="AC_000091_163"/>
                    </assert_contents>
                </element>
            </output_collection>
            <output name="concatenated_fasta" ftype="fasta">
                <assert_contents>
                    <has_text text="637000110"/>
                    <has_text text="MLKAGVHFGHQTRYW"/>
                </assert_contents>
            </output>
            <output name="concatenated_tre" ftype="phyloxml">
                <assert_contents>
                    <has_text text="IMG_646564547"/>
                    <has_text text="g__Methanocaldococcus"/>
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[
@HELP_HEADER@

This command places bins in the genome tree.

The following heuristic is used to establish the translation table used by Prodigal: use table 11 unless the coding density using table 4 is 5% higher than when using table 11 and the coding density under table 4 is >70%. Distinguishing between tables 4 and 25 is challenging so CheckM does not attempt to distinguish between these two tables. If you know the correct translation table for your genomes, it is recommended that you call genes outside of CheckM and provide CheckM with the protein sequences

    ]]></help>
    <expand macro="citations"/>
</tool>