<tool id="hybpiper" name="HybPiper" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="@PROFILE@">
    <description>Analyse targeted sequence capture data</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="xrefs"/>
    <expand macro="requirements"/>
    <command detect_errors="exit_code"><![CDATA[

    ## sample name checking
    #import re
    #def check_sample_name($sample_name):
        #if re.search(r'[^A-Za-z0-9_\-]', $sample_name):
            printf '%s\n'
            'ERROR: special characters detected in sample identifier.'
            'Identifiers may only contain letters, numbers, underscores and hyphens.'
            'Check the identifier for the following sample:' 
            '${sample_name}'
            1>&2
            &&
            exit 1
            &&
        #end if
    #end def

    ## set up files
    ln -s '${targetfile_dna}' ./target_file.fasta
    &&

    ###############################
    ## hybpiper check_targetfile ##
    ###############################

    #if str( $job_conditional.hybpiper_job ) == "check_and_fix_targetfile":
        hybpiper check_targetfile 
        --targetfile_dna target_file.fasta
        &&

        mv fix_targetfile*.ctl hybpiper.ctl
        &&

        hybpiper fix_targetfile
        --targetfile_dna target_file.fasta
        --allow_gene_removal
        hybpiper.ctl
        &&
    
    #######################
    ## hybpiper assemble ##
    #######################

    #elif str( $job_conditional.hybpiper_job ) == "assemble":
        #set sample_prefix = str($job_conditional.paired_input.element_identifier)

        $check_sample_name($sample_prefix)

        hybpiper assemble
        --readfiles
        '${job_conditional.paired_input.forward}'
        '${job_conditional.paired_input.reverse}'
        --targetfile_dna target_file.fasta
        --diamond
        --cpu \${GALAXY_SLOTS:-1}
        --prefix '${sample_prefix}'
        &&

        tar -cvf '${hybpiper_archive}' --directory='${sample_prefix}' .
        &&

    #######################################
    ## hybpiper stats/retrieve_sequences ##
    #######################################

    #elif str( $job_conditional.hybpiper_job ) == "stats":

        ## check logic of requested items
        #unless $job_conditional.stats_type_select or $job_conditional.sequence_type_select:
            printf '%s\n'
            'ERROR: No outputs selected.'
            1>&2
            &&
            exit 1
            &&
        #end unless
        #if $job_conditional.heatmap and not $job_conditional.stats_type_select:
            printf '%s\n'
            'ERROR: heatmap requested, but no stats selected.'
            1>&2
            &&
            exit 1
            &&
        #end if

        #for $sample in $job_conditional.hybpiper_results
            #set sample_prefix = str($sample.element_identifier)

            $check_sample_name($sample_prefix)

            mkdir -p '${sample_prefix}'
            &&

            tar -xf '${sample}' -C '${sample_prefix}'
            &&

            echo '${sample_prefix}' >> namelist.txt
            &&
        #end for

        ## Produce a stats file for each requested output type
        #for $stats_output in $job_conditional.stats_type_select:
            hybpiper stats
            --targetfile_dna target_file.fasta
            --stats_filename 'stats.${stats_output}'
            --seq_lengths_filename 'seq_lengths.${stats_output}'
            '${stats_output}'
            namelist.txt
            &&

            ## Produce heatmaps if selected
            #if $job_conditional.heatmap:
                hybpiper recovery_heatmap
                --heatmap_filename 'heatmap.${stats_output}'
                --heatmap_filetype svg
                'seq_lengths.${stats_output}.tsv'
                &&
            #end if
        #end for

        ## Produce sequences for each requested type
        #for $sequence_output in $job_conditional.sequence_type_select:
            mkdir 'fasta.${sequence_output}'
            &&
            hybpiper retrieve_sequences
            --targetfile_dna target_file.fasta
            --sample_names namelist.txt
            --fasta_dir 'fasta.${sequence_output}'
            '${sequence_output}'
            &&
        #end for
    #end if
    
    wait

]]></command>
    
    <inputs>
        <param argument="--targetfile_dna" type="data" format="fasta" label="Target file" help="Target file in FASTA format" />
    
        <conditional name="job_conditional">
            <param name="hybpiper_job" type="select" label="Type of hybpiper run">
                <option value="check_and_fix_targetfile">Check and fix targetfile</option>
                <option value="assemble" selected="true">Assemble target loci</option>
                <option value="stats">Extract sequences and/or stats from Hybpiper runs</option>
            </param>
    
            <when value="check_and_fix_targetfile"/>
    
            <when value="assemble">
                <param name="paired_input" format="fastqsanger" type="data_collection" collection_type="paired" label="Input reads" help="Your reads must be in a paired collection. See below for more information." />
            </when>
    
            <when value="stats">
                <param name="hybpiper_results" type="data_collection" collection_type="list" format="tar" multiple="true" label="Results from Hybpiper assemble runs" />
                <param name="stats_type_select" type="select" label="Choose statistics to report" display="checkboxes" multiple="true" optional="true">
                    <option value="gene" selected="true">Gene</option>
                    <option value="supercontig">Supercontig</option>
                </param>
                <param name="heatmap" type="boolean" checked="false" label="Produce a heatmap for each of the selected statistics" />
                <param name="sequence_type_select" type="select" display="checkboxes" label="Choose sequences to extract" multiple="true" optional="true">
                    <option value="dna" selected="true">DNA</option>
                    <option value="aa">Amino acid</option>
                    <option value="intron">Intron</option> 
                    <option value="supercontig">Supercontig</option>
                </param>
            </when>
        </conditional>
    </inputs>
    
    <outputs>
        <!-- check_targetfile output -->
        <data name="fixed_targetfile" label="${targetfile_dna.element_identifier} (fixed)" format="fasta" from_work_dir="target_file_fixed.fasta">
            <filter>job_conditional['hybpiper_job'] == 'check_and_fix_targetfile'</filter>
        </data>
        <collection type="list" name="output_targetfile" label="Hybpiper logs for ${targetfile_dna.element_identifier}">
            <data name="targetfile_ctl_file" label="Hybpiper .ctl file for ${on_string}" format="txt" from_work_dir="hybpiper.ctl" />
            <data name="targetfile_report" label="Hybpiper targetfile report" format="tabular" from_work_dir="fix_targetfile_report.tsv" />
            <filter>job_conditional['hybpiper_job'] == 'check_and_fix_targetfile'</filter>
        </collection>

        <!-- assemble output -->
        <data name="hybpiper_archive" format="tar">
            <filter>job_conditional['hybpiper_job'] == 'assemble'</filter>
        </data>

        <!-- stats / stats output -->
        <collection name="hybpiper_stats" type="list" label="Hybpiper statistics">
            <data name="stats_gene" label="Hybpiper statistics (gene)" format="tabular" from_work_dir="stats.gene.tsv">
                <actions>
                    <action name="column_names" type="metadata" default="Name,NumReads,ReadsMapped,PctOnTarget,GenesMapped,GenesWithContigs,GenesWithSeqs,GenesAt25pct,GenesAt50pct,GenesAt75pct,GenesAt150pct,ParalogWarningsLong,ParalogWarningsDepth,GenesWithoutStitchedContigs,GenesWithStitchedContigs,GenesWithStitchedContigsSkipped,GenesWithChimeraWarning,TotalBasesRecovered" />
                </actions>
            </data>
            <data name="stats_supercontig" label="Hybpiper statistics (supercontig)" format="tabular" from_work_dir="stats.supercontig.tsv">
                <actions>
                    <action name="column_names" type="metadata" default="Name,NumReads,ReadsMapped,PctOnTarget,GenesMapped,GenesWithContigs,GenesWithSeqs,GenesAt25pct,GenesAt50pct,GenesAt75pct,GenesAt150pct,ParalogWarningsLong,ParalogWarningsDepth,GenesWithoutStitchedContigs,GenesWithStitchedContigs,GenesWithStitchedContigsSkipped,GenesWithChimeraWarning,TotalBasesRecovered" />
                </actions>
            </data>
            <data name="seqlengths_gene" label="Assembled sequence lengths (gene)" format="tabular" from_work_dir="seq_lengths.gene.tsv"/>
            <data name="seqlengths_supercontig" label="Assembled sequence lengths (supercontig)" format="tabular" from_work_dir="seq_lengths.supercontig.tsv">
            </data>
            <filter>job_conditional['hybpiper_job'] == 'stats' and ('gene' in job_conditional['stats_type_select'] or 'supercontig' in job_conditional['stats_type_select'])</filter>
        </collection>

        <!-- stats/heatmap output -->
        <collection name="hybpiper_heatmaps" type="list" label="Hybpiper heatmaps">
            <discover_datasets pattern="heatmap\.(?P&lt;designation&gt;.+)\.svg" format="svg" recurse="false" />
            <filter>job_conditional['hybpiper_job'] == 'stats' and job_conditional['heatmap'] and job_conditional['heatmap'] is true</filter>            
        </collection>

        <!-- stats/sequences output -->         
        <collection name="dna_sequences" type="list" label="DNA sequences">
            <discover_datasets pattern="(?P&lt;designation&gt;.+)\.FNA" format="fasta" directory="fasta.dna" recurse="false" />
            <filter>job_conditional['hybpiper_job'] == 'stats' and 'dna' in job_conditional['sequence_type_select']</filter>
        </collection>
        <collection name="aa_sequences" type="list" label="Amino acid sequences">
            <discover_datasets pattern="(?P&lt;designation&gt;.+)\.FAA" format="fasta" directory="fasta.aa" recurse="false" />
            <filter>job_conditional['hybpiper_job'] == 'stats' and 'aa' in job_conditional['sequence_type_select']</filter>
        </collection>
        <collection name="intron_sequences" type="list" label="Intron sequences">
            <discover_datasets pattern="(?P&lt;designation&gt;.+)\.fasta" format="fasta" directory="fasta.intron" recurse="false" />
            <filter>job_conditional['hybpiper_job'] == 'stats' and 'intron' in job_conditional['sequence_type_select']</filter>
        </collection>
        <collection name="supercontig_sequences" type="list" label="Supercontig sequences">
            <discover_datasets pattern="(?P&lt;designation&gt;.+)\.fasta" format="fasta" directory="fasta.supercontig" recurse="false" />
            <filter>job_conditional['hybpiper_job'] == 'stats' and 'supercontig' in job_conditional['sequence_type_select']</filter>
        </collection>

        <!-- dummy output, in case the user deselects everything -->
        <data name="dummy_output" label="Stats or sequences from Hybpiper runs" from_work_dir="namelist.txt" format="txt">
            <filter>job_conditional['hybpiper_job'] == 'stats' and not (job_conditional['stats_type_select'] or job_conditional['sequence_type_select']) </filter>
        </data>

    </outputs>
    <tests>

    <!-- test1: check and fix targetfile -->
    <test expect_num_outputs="4">
        <param name="targetfile_dna" value="test_targets.fasta.gz"/>
        <conditional name="job_conditional">
            <param name="hybpiper_job" value="check_and_fix_targetfile"/>
        </conditional>
        <output name="fixed_targetfile" file="test1_out.fasta"/>
        <output_collection name="output_targetfile" type="list" count="2">
            <element name="targetfile_ctl_file" file="test1_out.ctl"/>    
            <element name="targetfile_report" file="test1_out.tsv"/>
        </output_collection>
    </test>

    <!-- test2: assemble with paired collection -->
    <!-- Not possible to test stats unless element_identifier can be set. -->
    <test expect_failure="true">
        <param name="targetfile_dna" value="test_targets.fasta.gz"/>
        <conditional name="job_conditional">
            <param name="hybpiper_job" value="assemble"/>
            <param name="paired_input">
                <collection type="paired">
                    <element name="forward" ftype="fastqsanger.gz" value="NZ874_R1_test.fastq.gz" />
                    <element name="reverse" ftype="fastqsanger.gz" value="NZ874_R2_test.fastq.gz" />
                </collection>
            </param>
        </conditional>
        <!-- <output name="hybpiper_archive">
            <assert_contents>
                <has_size value="2386944" delta="200000" />
            </assert_contents>
        </output> -->
    </test>

    <!-- test3: all stats output -->
    <test expect_num_outputs="10">
        <param name="targetfile_dna" value="test_targets.fasta.gz"/>
        <conditional name="job_conditional">
            <param name="hybpiper_job" value="stats"/>
            <param name="hybpiper_results" >
                <collection type="list">
                    <element name="NZ874" value="NZ874.tar.gz" />
                </collection>
            </param>
            <param name="stats_type_select" value="gene,supercontig"/>
            <param name="heatmap" value="true"/>
            <param name="sequence_type_select" value="dna,aa,intron,supercontig"/>
        </conditional>
        <output_collection name="hybpiper_stats" type="list" count="4" />
        <output_collection name="hybpiper_heatmaps" type="list" count="2">
        </output_collection>
        <output_collection name="dna_sequences" type="list" count="13">
        </output_collection>
        <output_collection name="aa_sequences" type="list" count="13">
        </output_collection>
        <output_collection name="intron_sequences" type="list" count="13">
        </output_collection>
        <output_collection name="supercontig_sequences" type="list" count="13">
        </output_collection>
    </test>

    <!-- test4: no output selected -->
    <test expect_failure="true">
        <param name="targetfile_dna" value="test_targets.fasta.gz"/>
        <conditional name="job_conditional">
            <param name="hybpiper_job" value="stats"/>
            <param name="hybpiper_results" >
                <collection type="list">
                    <element name="NZ874" value="NZ874.tar.gz" />
                </collection>
            </param>
            <param name="stats_type_select" value=""/>
            <param name="heatmap" value="true"/>
            <param name="sequence_type_select" value=""/>
        </conditional>
    </test>

</tests>
    <help><![CDATA[

Using HybPiper on Galaxy
------------------------

Input
~~~~~

On Galaxy, **you have to use paired collections as input** for
HybPiper assemblies. HybPiper relies on the directory hierarchy it creates for each
sample during assembly. The hierarchy is based on the name of the
sample, which you provide to Galaxy as the identifier in the collection.

Using paired collections
~~~~~~~~~~~~~~~~~~~~~~~~

If you have your sequencing reads in individual datasets, you can easily organise them into a paired
collection. See the Galaxy training material on `using dataset
collections <https://gxy.io/GTN:T00146>`__
for a step-by-step guide.

**Note**: because HybPiper uses sample
identifiers to create directories, you **can't use special characters**
in your sample identifiers. The only allowed characters are letters,
numbers, underscores and hyphens.

You can't use single-end and unpaired reads as input to Hybpiper on Galaxy.

Running HybPiper
~~~~~~~~~~~~~~~~

The following HybPiper analyses are available on Galaxy:

1. Check your target file and fix issues (optional)
2. Assemble target loci per-sample
3. Extract sequences and summary statistics

Use the *Type of hybpiper run* drop-down to select an analysis.

.. class:: infomark

What it does
------------

HybPiper was designed for processing targeted sequence capture data. In
targeted sequence capture, DNA sequencing libraries are enriched for
gene regions of interest. This is used for sequencing many loci
simultaneously based on bait sequences.

HybPiper is a suite of scripts that wrap and connect other tools to
extract target sequences from the sequencing reads. The HybPiper
pipeline starts with high-throughput sequencing reads (for example from
Illumina MiSeq), and assigns them to target genes using DIAMOND. The
reads are distributed to separate directories, where they are assembled
separately using SPAdes. The main output is a collection of FASTA files
of the (in frame) CDS portion of the sample for each target region. You
can also generate a separate collections of files with the translated
protein sequences, the intronic regions flanking each exon, and putative
paralog sequences.

For more information, please see `the HybPiper
wiki <https://github.com/mossmatters/HybPiper/wiki>`__.


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