comparison ncbi_egapx.xml @ 3:4420dd857c41 draft

planemo upload for repository https://github.com/richard-burhans/galaxytools/tree/main/tools/ncbi_egapx commit 050f870384e004445d3dcfb56302b7894793bd23

2:7c72d5c7e449

    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="edam_ontology"/>
    <expand macro="requirements"/>
    <command detect_errors="exit_code"><![CDATA[
    source /galaxy/env.bash &&
    echo \${PATH} &&
    ln -s /galaxy/egapx/egapx_config &&
    python3 /galaxy/egapx/ui/egapx.py '$yamlconfig' -e galaxy -o 'egapx_out'
    ]]></command>
    <inputs>
      <param name="yamlconfig" type="data" optional="false" label="egapx configuration yaml file to execute" help="" format="yaml,txt" multiple="false"/>
    </inputs>
    <outputs>
      <collection name="egapx_out" type="list" label="Outputs from egapx">
        <discover_datasets pattern="__name_and_ext__" directory="egapx_out" visible="false"/>
      </collection>
    </outputs>
    <tests>
      <test expect_test_failure="true">
        <param name="yamlconfig" value="input.yaml"/>
        <output_collection name="egapx_out" type="list" count="8"/>
      </test>
    </tests>
  <help><![CDATA[
Galaxy tool wrapping the Eukaryotic Genome Annotation Pipeline (EGAPx)
=================================================================================================

.. class:: warningmark

**Proof of concept: a quick hack to run a NF workflow inside a specialised Galaxy tool wrapper**

EGAPx is a big, complicated Nextflow workflow, challenging and costly to re-implement **properly**, requiring dozens of new tools and replicating a lot of
complicated *groovy* workflow logic.

It is also very new and in rapid development. Investing developer effort and keeping updated as EGAPx changes rapidly may be *inefficient of developer resources*.

This wrapper is designed to allow measuring how *inefficient* it is in terms of computing resource utilisation, in comparison to the developer effort
required to convert Nextflow DDL into tools and WF logic. Balancing these competing requirements is a fundamental Galaxy challenge.


EGAPx requires very substantial resources to run with real data. *128GB and 32 cores* are the minimum requirement; *256GB and 64 cores* are recommended.

A special minimal example that can be run in 6GB with 4 cores is provided as a yaml configuration and is used for the tool test.

In this implementation, the user must supply a yaml configuration file as initial proof of concept.
History inputs and even a yaml editor might be provided in future.

3:4420dd857c41

    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="edam_ontology"/>
    <expand macro="requirements"/>
    <command detect_errors="aggressive"><![CDATA[
    #if str($cond_input_style.input_style) == "history":
      #set yamlconfig = $yamlin
    #else:
      #set yamlconfig = 'egapx.yaml'
      rm -rf 'egapx.yaml' &&
      touch 'egapx.yaml' &&
      echo '# yaml generated by ncbi_egapx.xml' >> egapx.yaml &&
      echo 'taxid:  $taxid' >> egapx.yaml &&
      #if str($reference_genome.genome_type_select) == "indexed":
        echo 'genome:  $reference_genome.genome.fields.path' >> 'egapx.yaml' &&
      #elif str($reference_genome.genome_type_select) == "history"
        echo 'genome:  $reference_genome.genome'  >> 'egapx.yaml' &&
      #else:
        echo 'genome:  $reference_genome.uri' >> 'egapx.yaml' &&
      #end if
      echo 'reads:' >> 'egapx.yaml' &&
      #if str($condrnaseq.rna_type_select) == "history":
        #for $r in $rnaseq:
          echo '  - $r'  >> 'egapx.yaml' &&
        #end for
      #else:
        #set rs = $rnaseq.split()
        #set rsplit = [x.strip() for x in $rs]
        #for $r in $rsplit:
          echo '  - $r'  >> 'egapx.yaml' &&
        #end for
      #end if
      #if len($xtra.strip()) > 0:
        #set lxtra = $xtra.split('\n')
        #for row in $lxtra:
            echo '$row' >> 'egapx.yaml' &&
        #end for
      #end if
      echo '' >> 'egapx.yaml' &&
      echo "Calculated contents of egapx yaml" &&
      cat 'egapx.yaml' &&
    #end if
    source /galaxy/env.bash &&
    echo \${PATH} &&
    ln -s /galaxy/egapx/egapx_config &&
    python3 /galaxy/egapx/ui/egapx.py '$yamlconfig' -e galaxy  -o 'egapx_out'
    ]]></command>
    <inputs>
    <conditional name="cond_input_style">
      <param name="input_style" type="select" label="Fill in a tool form or use an existing yaml configuration from the current history?"
        help="Use a pre-prepared yaml if available. Use the tool form if history files are needed as rna-seq or reference genome inputs for this job">
        <option selected="True" value="history">Use a pre-prepared yaml egapx configuration</option>
        <option value="fillform">Provide configuration details for conversion into a configuration yaml</option>
      </param>
      <when value="history">
        <param name="yamlin" type="data" optional="false" label="egapx configuration yaml file to pass to Nextflow" help="" format="yaml,txt"/>
      </when>
      <when value="fillform">
          <param name="taxid" type="text" optional="false" label="NCBI Taxon ID" help="Used to identify the HMM model files needed"/>
          <conditional name="reference_genome">
            <param name="genome_type_select" type="select" label="Reference genome source for mapping supplied RNA-seq reads" 
              help="Select a built in, history or remote URI for the reference genome fasta">
                <option value="indexed">Use a Galaxy server built-in genome</option>
                <option value="history" selected="True">Use a genome fasta file from the current history</option>
                <option value="uri">Provide a remote web link URI ("https://...") pointing at the required genome reference fasta file</option>
            </param>
            <when value="indexed">
                <param name="genome" type="select" optional="false" label="Select a built in reference genome or custom genome" 
                  help="If not listed, add a custom genome or use a reference genome from the history">
                    <options from_data_table="all_fasta">
                        <validator message="No genomes are available " type="no_options"/>
                    </options>
                </param>
            </when>
            <when value="history">
                <param name="genome" type="data" format="fasta" optional="false" label="Select the reference genome fasta from the current history"/>
            </when>
            <when value="uri">
                <param name="uri" type="text" label="URI pointing to the reference genome fasta file" help=""/>
            </when>
          </conditional>
          <conditional name="condrnaseq">
            <param name="rna_type_select" type="select" label="RNA sequence data source" help="Select RNAseq input data from history or input a list of SRA identifiers or remote URI">
                <option selected="True" value="list">Type in a list of SRA identifiers and/or remote RNA-seq fasta URI</option>
                <option value="history">Select one or more RNA-seq fastq datasets from the current history</option>
            </param>
            <when value="history">
              <param name="rnaseq" type="data" format="fastqsanger, fastqsanger.gz" optional="false" multiple="true" 
                label="Select multiple RNA-seq fastqsanger inputs from the current history" help="All selected rna-seq fastqsanger will be added to the yaml for egapx configuration"/>
            </when>
            <when value="list">
                <param name="rnaseq" type="text" area="true" optional="false" label="List all required individual RNA-seq URI or SRA identifiers, separated by spaces or newlines"
                    help="Either a working URI for a RNA-seq fasta, or a bare SRA identifier will work - can be mixed">
                    <validator type="empty_field"/>
                </param>
            </when>
          </conditional>
          <param name="xtra" type="text" area="true" optional="true" label="Additional yaml to append to the egapx.yaml configuration"
                    help="Not normally needed but useful for testing additional configuration elements">
            <sanitizer invalid_char="">
                  <valid initial="string.printable">
                </valid>
            </sanitizer>
          </param>
      </when>
    </conditional>
    </inputs>
    <outputs>
      <collection name="egapx_out" type="list" label="Outputs from egapx">
        <discover_datasets pattern="__name_and_ext__" directory="egapx_out" visible="false"/>
      </collection>
    </outputs>
    <tests>
      <test expect_test_failure="true">
        <param name="yamlin" value="input.yaml"/>
        <output_collection name="egapx_out" type="list" count="8"/>
      </test>
    </tests>
  <help><![CDATA[
Galaxy tool wrapping the Eukaryotic Genome Annotation Pipeline (EGAPx)
=================================================================================================

.. class:: warningmark

**Proof of concept: a hack to run a NF workflow inside a specialised Galaxy tool wrapper**

EGAPx is a big, complicated Nextflow workflow, challenging and costly to re-implement **properly**, requiring dozens of new tools and replicating a lot of
complicated *groovy* workflow logic.

It is also very new and in rapid development. Investing developer effort and keeping updated as EGAPx changes rapidly may be *inefficient of developer resources*.

This wrapper is designed to allow measuring how *inefficient* it is in terms of computing resource utilisation, in comparison to the developer effort
required to convert Nextflow DDL into tools and WF logic. Balancing these competing requirements is a fundamental Galaxy challenge.


EGAPx requires very substantial resources to run with real data. *132GB and 32 cores* are the minimum requirement; *256GB and 64 cores* are recommended.

A special minimal example that can be run in 6GB with 4 cores is provided as a yaml configuration and is used for the tool test.

In this implementation, the user must supply a yaml configuration file as initial proof of concept.
History inputs and even a yaml editor might be provided in future.