annotate run_sccaf.xml @ 0:ca26d4b4b02c draft

planemo upload for repository https://github.com/ebi-gene-expression-group/container-galaxy-sc-tertiary/tree/develop/tools/tertiary-analysis/sccaf commit 68be7a6fdb93e8b59e80e5f16e7fecdaa16f288c

<?xml version="1.0" encoding="utf-8"?>
<tool id="run_sccaf" name="Run SCCAF" version="@TOOL_VERSION@+galaxy0">
  <description>to assess and optimise clustering</description>
  <macros>
    <import>sccaf_macros.xml</import>
  </macros>
  <expand macro="requirements"/>
  <command detect_errors="exit_code"><![CDATA[
ln -s ${input_obj_file} input.h5 &&
sccaf -i input.h5

--cores \${GALAXY_SLOTS:-1}

#if $cluster_source.use_tsv
  --external-clustering-tsv ${cluster_source.input_tsv}
#else
  --slot-for-existing-clustering ${cluster_source.obj_attr}
#end if

#if $mode.skip_init_assessment
  --skip-assessment
#end if

#if $mode.optimise
  #if not $mode.init.from_input
  --resolution ${mode.init.resolution}
  #end if
  --optimise
  #if $mode.optimisation_stop.condition == "accuracy_threshold"
    --min-accuracy ${mode.optimisation_stop.accuracy}
  #else
    --min-accuracy 0.955
    --undercluster-boundary ${mode.optimisation_stop.min_resolution_clustering_slot}
  #end if
  --produce-rounds-summary

  --optimisation-plots-output ${opt_pdf_out}

  #if $mode.citer
    --conf-sampling-iterations ${mode.citer}
  #end if

  > ${opt_text_out};
#end if

#if $mode.optimise and $mode.distribute_assesment:
mkdir -p outputs;
for round in \$(cat rounds.txt); do
  echo "Round: "\$round;
  echo \$round > outputs/round_\$round\.txt;
done
#end if



]]></command>

  <inputs>
    <expand macro="input_object_params"/>

    <conditional name="cluster_source">
      <param name="use_tsv" type="boolean" checked="true" label="Use external cluster information" help="If the provided AnnData/Loom file does not include the clustering, or if you want to use an external clustering assigment."/>
      <when value="true">
        <param name="input_tsv" type="data" argument="--external-clustering-tsv" format="tsv" label="Cluster table for assessment in tsv format"/>
      </when>
      <when value="false">
        <param name="obj_attr" type="text" argument="--slot-for-existing-clustering" value="louvain" label="Attribute in input object that contains cluster information" help="If you are not using an external clustering, then you must specify the slot/index in the AnnData or Loom file where the clustering to be used for assessment (and potentially optimisation starting point) is saved."/>
      </when>
    </conditional>

    <conditional name="mode">
      <param name="optimise" type="boolean" checked="false" argument="--optimise" label="Run clustering optimisation" help="By default the tool only runs an assesment of the clustering quality. To further optimise the clustering, enable this option."/>
      <when value="true">
        <conditional name="init">
          <param name="from_input" type="boolean" checked="true" label="Use input clustering to initialise optimisation" help="This option uses the previously specified clustering (either external file provided or internal index/slot specified) to serve as starting point for the optimisation process."/>
          <when value="true">
          </when>
          <when value="false">
            <param name="resolution" type="float" value="1.5" argument="--resolution" label="Resolution for initialising louvain clustering" help="Perform an initial clustering and use this as an starting point."/>
          </when>
        </conditional>
        <conditional name="optimisation_stop">
          <param type="select" name="condition" help="How should the optimisation algorithm be stopped." label="Stop condition">
            <option value="accuracy_threshold" selected="true">Set a threshold for accuracy</option>
            <option value="min_resolution_clustering">Use a slot with a low resolution clustering</option>
          </param>
          <when value="accuracy_threshold">
            <param name="accuracy" type="float" value="0.90" argument="--accuracy" label="Accuracy for convergence of the optimisation process" help="The SCCAF optimisation process will converge once this accuracy is achieved."/>
          </when>
          <when value="min_resolution_clustering">
            <param name="min_resolution_clustering_slot" type="text" argument="--undercluster-boundary" label="Underclustering boundary to use in the optimisation." help="The slot inside the ann data object with the desired underclustering (low resolution) to be used as exit condition."/>
          </when>
        </conditional>
        <param name="skip_init_assessment" type="boolean" argument="--skip-assessment" checked="false" label="Run only the optimisation and skip the initial assessment" help="If you are running the optimisation, you can choose skip the initial assessment to save time."/>
        <param name="distribute_assesment" type="boolean" checked="false" label="Produce parameter walk for asessment distribution" help="Will split rounds in files so that multiple assess processes can be distributed down the line"/>
        <param name="citer" type="integer" value="3" argument="--conf-sampling-iterations" label="Number of iterations of sampling for the confusion matrix." help="By default is 3, higher numbers mean more stable results (number of rounds) but also longer execution times." optional="True"/>
      </when>
      <when value="false">
      </when>
    </conditional>
  </inputs>

  <outputs>
    <data name="output_png" format="png" from_work_dir="roc-curve.png" label="${tool.name} on ${on_string} ROC Curve for initial assesment">
      <filter>not mode['skip_init_assessment']</filter>
    </data>
    <data name="output_h5" format="h5" from_work_dir="output.h5" label="${tool.name} on ${on_string} output.h5">
      <filter>mode['optimise']</filter>
    </data>
    <data name="optim_png" format="png" from_work_dir="optim.png" label="${tool.name} on ${on_string} ROC Curve for optimisation result">
      <filter>mode['optimise']</filter>
    </data>
    <collection name="rounds_walk" type="list" label="Rounds for assesment distribution">
      <filter>mode['optimise'] and mode['distribute_assesment']</filter>
      <discover_datasets pattern="__name_and_ext__" directory="outputs"/>
    </collection>
    <data name="opt_text_out" format="txt" label="${tool.name} on ${on_string} Optimisation process log">
      <filter>mode['optimise']</filter>
    </data>
    <data name="opt_pdf_out" format="pdf" label="${tool.name} on ${on_string} Optimisation process plots">
      <filter>mode['optimise']</filter>
    </data>
  </outputs>

  <tests>
    <test>
      <param name="input_obj_file" value="find_cluster.h5"/>
      <param name="use_tsv" value="true"/>
      <param name="input_tsv" value="find_cluster.tsv"/>
      <output name="output_png" file="run_sccaf.png" ftype="png"/>
    </test>
  </tests>

  <help><![CDATA[
@SCCAF_INTRO@

Inputs
------

* AnnData object which contains the expression matrix and pre-calculated coordinates for UMAP. The AnnData object can include already clustering data, in which case the user will need to know on which AnnData slot/label is contained.
* Optional external text file with mappings between cells and clusters (when no clustering is given inside the AnnData file).

Modes of operation
------------------

* Optimisation with exit condition based on accuracy cut-off. In this case the user provides a minimum cut-off for the accuracy to be achieved and the optimisation process will exit at that point.
* Optimisation with exit condition based on under-clustered scenario. In the case the AnnData object given must include a low resolution, under clustered clustering, and its label must be know to be specified.
* Assesment only (no optimisation), where an existing clustering is assessed.

This is resource intensive.

Distributed assesment
---------------------

If running the optimisation you can distribute assessments of the optimisation
results. For this, activate the "Produce parameter walk for assessment
distribution" option, which will generate a "Rounds for assesment distribution".
Then feed the AnnData output of the optimisation process and the rounds output
to the SCCAF Assesment module. Then merge all assessment results with
SCCAF Assesment Merger (this also receives the rounds output). The workflow
would look like this:

.. image:: example_sccaf_workflow.png
   :height: 400 px
   :width: 850 px
   :scale: 80 %


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