annotate manipulate.xml @ 2:a56baceb1900 draft

"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/anndata/ commit 3b41d687ff30583540d055f6995de00530cca81d"

<tool id="anndata_manipulate" name="Manipulate AnnData" version="@VERSION@+@GALAXY_VERSION@">
    <description>object</description>
    <macros>
        <import>macros.xml</import>
        <xml name="param_join">
            <param name="join" type="select" label="The connecting string between name and integer">
                <option value="-">-</option>
                <option value="_">_</option>
                <option value=" "> </option>
                <option value="/">/</option>
            </param>
        </xml>
    </macros>
    <expand macro="requirements"/>
    <expand macro="version_command"/>
    <command detect_errors="exit_code"><![CDATA[
@CMD@
      ]]></command>
    <configfiles>
        <configfile name="script_file"><![CDATA[
@CMD_imports@

adata = ad.read('$input')

#if $manipulate.function == 'concatenate'
    #for i, filepath in enumerate($manipulate.other_adatas)
adata_$i = ad.read('$filepath')
    #end for
adata.concatenate(
    #for i, filepath in enumerate($manipulate.other_adatas)
    adata_$i,
    #end for
    join='$manipulate.join',
    #if str($manipulate.index_unique) != ''
    index_unique='$manipulate.index_unique',
    #else
    index_unique=None,
    #end if
    batch_key='$manipulate.batch_key')

#else if $manipulate.function == 'var_names_make_unique'
adata.var_names_make_unique(join='$manipulate.join')

#else if $manipulate.function == 'obs_names_make_unique'
adata.obs_names_make_unique(join='$manipulate.join')

#else if $manipulate.function == 'rename_categories'
    #set $categories = [x.strip() for x in str($manipulate.categories).split(',')]
adata.rename_categories(
    key='$manipulate.key',
    categories=$categories)

#else if $manipulate.function == 'strings_to_categoricals'
adata.strings_to_categoricals()

#else if $manipulate.function == 'transpose'
adata = adata.transpose()

#else if $manipulate.function == 'add_annotation'
import pandas as pd
extra_annot_t = pd.read_csv('$manipulate.new_annot', sep='\t').reset_index(drop=True)
    #if $manipulate.var_obs == 'var'
var_index = adata.var_names
var = pd.concat([adata.var.reset_index(drop=True), extra_annot_t], axis=1)
var.index = var_index
adata.var = var
    #else if $manipulate.var_obs == 'obs'
obs_index = adata.obs.index
obs = pd.concat([adata.obs.reset_index(drop=True), extra_annot_t], axis=1)
obs.index = obs_index
adata.obs = obs
    #end if

#else if $manipulate.function == 'filter'
    #if $manipulate.filter.filter == 'key'
        #if $manipulate.var_obs == 'var'
filtered = adata.var['$manipulate.filter.key']
        #else if $manipulate.var_obs == 'obs'
filtered = adata.obs['$manipulate.filter.key']
        #end if

        #if $manipulate.filter.filter_key.type == 'number'
            #if $manipulate.filter.filter_key.filter == 'equal'
filtered = filtered == $manipulate.filter.filter_key.value
            #else if $manipulate.filter.filter_key.filter == 'equal'
filtered = filtered != $manipulate.filter.filter_key.value
            #else if $manipulate.filter.filter_key.filter == 'less'
filtered = filtered < $manipulate.filter.filter_key.value
            #else if $manipulate.filter.filter_key.filter == 'less_or_equal'
filtered = filtered <= $manipulate.filter.filter_key.value
            #else if $manipulate.filter.filter_key.filter == 'greater'
filtered = filtered > $manipulate.filter.filter_key.value
            #else if $manipulate.filter.filter_key.filter == 'greater_or_equal'
filtered = filtered >= $manipulate.filter.filter_key.value
            #end if
        #else if $manipulate.filter.filter_key.type == 'text'
            #if $manipulate.filter.filter_key.filter == 'equal'
filtered = filtered == '$manipulate.filter.filter_key.value'
            #else
filtered = filtered != '$manipulate.filter.filter_key.value'
            #end if
        #else if $manipulate.filter.filter_key.type == 'boolean'
filtered = filtered == $manipulate.filter.filter_key.value
        #end if

    #else if $manipulate.filter.filter == 'index'
        #if str($manipulate.filter.index.format) == 'file'
with open('$manipulate.filter.index.file', 'r') as filter_f:
    filters = [str(x.strip()) for x in filter_f.readlines()]
filtered = filters
        #else
            #set $filters = [str(x.strip()) for x in $manipulate.filter.index.text.split(',')]
filtered = $filters
        #end if
    #end if
print(filtered)

    #if $manipulate.var_obs == 'var'
adata = adata[:,filtered]
    #else if $manipulate.var_obs == 'obs'
adata = adata[filtered, :]
    #end if

#else if $manipulate.function == 'save_raw'
adata.raw = adata

#end if

adata.write('anndata.h5ad')
]]></configfile>
    </configfiles>
    <inputs>
        <param name="input" type="data" format="h5ad" label="Annotated data matrix"/>
        <conditional name="manipulate">
            <param name="function" type="select" label="Function of manipulate the object">
                <option value="concatenate">Concatenate along the observations axis</option>
                <option value="obs_names_make_unique">Makes the obs index unique by appending '1', '2', etc</option>
                <option value="var_names_make_unique">Makes the var index unique by appending '1', '2', etc</option>
                <option value="rename_categories">Rename categories of annotation</option>
                <option value="strings_to_categoricals">Transform string annotations to categoricals</option>
                <option value="transpose">Transpose the data matrix, leaving observations and variables interchanged</option>
                <option value="add_annotation">Add new annotation(s) for observations or variables</option>
                <option value="filter">Filter observations or variables</option>
                <option value="save_raw">Freeze the current state into the 'raw' attribute</option>
            </param>
            <when value="concatenate">
                <param name="other_adatas" type="data" format="h5ad" multiple="true" label="Annotated data matrix to add"/>
                <param name="join" type="select" label="Join method">
                    <option value="inner">Intersection of variables</option>
                    <option value="outer">Union of variables</option>
                </param>
                <param name="batch_key" type="text" value="batch" label="Key to add the batch annotation to obs"/>
                <param name="index_unique" type="select" label="Separator to join the existing index names with the batch category" help="Leave it empty to keep existing indices">
                    <option value="-">-</option>
                    <option value="_">_</option>
                    <option value=" "> </option>
                    <option value="/">/</option>
                </param>
            </when>
            <when value="obs_names_make_unique">
                <expand macro="param_join"/>
            </when>
            <when value="var_names_make_unique">
                <expand macro="param_join"/>
            </when>
            <when value="rename_categories">
                <param name="key" type="text" value="" label="Key for observations or variables annotation" help="Annotation key in obs or var"/>
                <param name="categories" type="text" value="" label="Comma-separated list of new categories" help="It should be the same number as the old categories"/>
            </when>
            <when value="strings_to_categoricals"/>
            <when value="transpose"/>
            <when value="add_annotation">
                <param name="var_obs" type="select" label="What to annotate?">
                    <option value="var">Variables (var)</option>
                    <option value="obs">Observations (obs)</option>
                </param>
                <param name="new_annot" type="data" format="tabular" label="Table with new annotations" 
                    help="The new table should have the same number of rows and same order than obs or var. The key names should be in the header (1st line)"/>
            </when>
            <when value="filter">
                <param name="var_obs" type="select" label="What to filter?">
                    <option value="var">Variables (var)</option>
                    <option value="obs">Observations (obs)</option>
                </param>
                <conditional name="filter">
                    <param name="filter" type="select" label="Type of filtering?">
                        <option value="key">By key (column) values</option>
                        <option value="index">By index (row)</option>
                    </param>
                    <when value="key">
                        <param name="key" type="text" value="n_genes" label="Key to filter"/>
                        <conditional name="filter_key">
                            <param name="type" type="select" label="Type of value to filter">
                                <option value="number">Number</option>
                                <option value="text">Text</option>
                                <option value="boolean">Boolean</option>
                            </param>
                            <when value="number">
                                <param name="filter" type="select" label="Filter">
                                    <option value="equal">equal to</option>
                                    <option value="not_equal">not equal to</option>
                                    <option value="less">less than</option>
                                    <option value="less_or_equal">less than or equal to</option>
                                    <option value="greater">greater than</option>
                                    <option value="greater_or_equal">greater than or equal to</option>
                                </param>
                                <param name="value" type="float" value="2500" label="Value"/>
                            </when>
                            <when value="text">
                                <param name="filter" type="select" label="Filter">
                                    <option value="equal">equal to</option>
                                    <option value="not_equal">not equal to</option></param>
                                <param name="value" type="text" value="2500" label="Value"/>
                            </when>
                            <when value="boolean">
                                <param name="value" type="boolean" truevalue="True" falsevalue="False" checked="true" label="Value to keep"/>
                            </when>
                        </conditional>
                    </when>
                    <when value="index">
                        <conditional name="index">
                            <param name="format" type="select" label="Format for the filter by index">
                                <option value="file">File</option>
                                <option value="text" selected="true">Text</option>
                            </param>
                            <when value="text">
                                <param name="text" type="text" value="" label="List of index to keep" help="Indexes separated by a comma"/>
                            </when>
                            <when value="file">
                                <param name="file" type="data" format="txt" label="File with the list of index to keep" help="One index per line"/>
                            </when>
                        </conditional>
                    </when>
                </conditional>
            </when>
        </conditional>
    </inputs>
    <outputs>
        <data name="anndata" format="h5ad" from_work_dir="anndata.h5ad" label="${tool.name} (${manipulate.function}) on ${on_string}"/>
    </outputs>
    <tests>
        <test>
            <!-- test 1 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="concatenate"/>
                <param name="other_adatas" value="import.csv.h5ad"/>
                <param name="join" value="inner"/>
                <param name="batch_key" value="batch"/>
                <param name="index_unique" value="-"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="adata_0"/>
                <has_text_matching expression="adata.concatenate"/>
                <has_text_matching expression="join='inner'"/>
                <has_text_matching expression="index_unique='-'"/>
                <has_text_matching expression="batch_key='batch'"/>
            </assert_stdout>
            <output name="anndata" value="manipulate.concatenate.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 2 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="obs_names_make_unique"/>
                <param name="join" value="-"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="adata.obs_names_make_unique\(join='-'\)"/>
            </assert_stdout>
            <output name="anndata" value="manipulate.obs_names_make_unique.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 3 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="var_names_make_unique"/>
                <param name="join" value="-"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="adata.var_names_make_unique\(join='-'\)"/>
            </assert_stdout>
            <output name="anndata" value="manipulate.var_names_make_unique.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 4 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="rename_categories"/>
                <param name="key" value="cell_type"/>
                <param name="categories" value="Er,Ml,Mt, Ne, pr"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="adata.rename_categories"/>
                <has_text_matching expression="key='cell_type'"/>
                <has_text_matching expression="categories=\['Er', 'Ml', 'Mt', 'Ne', 'pr'\]"/>
            </assert_stdout>
            <output name="anndata" value="manipulate.rename_categories.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 5 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="strings_to_categoricals"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="adata.strings_to_categoricals"/>
            </assert_stdout>
            <output name="anndata" value="manipulate.strings_to_categoricals.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 6 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="transpose"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="adata.transpose"/>
            </assert_stdout>
            <output name="anndata" value="manipulate.transpose.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 7 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="add_annotation"/>
                <param name="var_obs" value="var"/>
                <param name="new_annot" value="var_add_annotation.tabular"/>
            </conditional>
            <output name="anndata" value="manipulate.add_annotation_var.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 8 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="add_annotation"/>
                <param name="var_obs" value="obs"/>
                <param name="new_annot" value="obs_add_annotation.tabular"/>
            </conditional>
            <output name="anndata" value="manipulate.add_annotation_obs.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 9 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="filter"/>
                <param name="var_obs" value="var"/>
                <conditional name="filter">
                    <param name="filter" value="index"/>
                    <conditional name="index">
                        <param name="format" value="text"/>
                        <param name="text" value="Gata2,EKLF"/>
                    </conditional>
                </conditional>
            </conditional>
            <output name="anndata" value="manipulate.filter_var_index.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 10 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="filter"/>
                <param name="var_obs" value="obs"/>
                <conditional name="filter">
                    <param name="filter" value="key"/>
                    <param name="key" value="cell_type"/>
                    <conditional name="filter_key">
                        <param name="type" value="text"/>
                        <param name="filter" value="equal"/>
                        <param name="value" value="progenitor"/>
                    </conditional>
                </conditional>
            </conditional>
            <output name="anndata" value="manipulate.filter_obs_key.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
        <test>
            <!-- test 11 -->
            <param name="input" value="krumsiek11.h5ad"/>
            <conditional name="manipulate">
                <param name="function" value="save_raw"/>
            </conditional>
            <output name="anndata" value="manipulate.save_raw.h5ad" ftype="h5ad" compare="sim_size"/>
        </test>
    </tests>
    <help><![CDATA[
**What it does**

This tool takes a AnnData dataset, manipulates it and returns it. 

The possible manipulations are: 

- Concatenate along the observations axis (`concatenate method <https://anndata.readthedocs.io/en/latest/anndata.AnnData.concatenate.html>`__) 

    The `uns`, `varm` and `obsm` attributes are ignored.

    If you use `join='outer'` this fills 0s for sparse data when variables are absent in a batch. Use this with care. Dense data is filled with `NaN`

- Makes the obs index unique by appending '1', '2', etc (`obs_names_make_unique method <https://anndata.readthedocs.io/en/latest/anndata.AnnData.obs_names_make_unique.html>`__) 

    The first occurance of a non-unique value is ignored.

- Makes the var index unique by appending '1', '2', etc (`var_names_make_unique method <https://anndata.readthedocs.io/en/latest/anndata.AnnData.var_names_make_unique.html>`__) 

    The first occurance of a non-unique value is ignored.

- Rename categories of annotation `key` in `obs`, `var` and `uns` (`rename_categories method <https://anndata.readthedocs.io/en/latest/anndata.AnnData.rename_categories.html>`__) 

    Besides calling `self.obs[key].cat.categories = categories` - similar for `var` - this also renames categories in unstructured annotation that uses the categorical annotation `key`

- Transform string annotations to categoricals (`strings_to_categoricals method <https://anndata.readthedocs.io/en/latest/anndata.AnnData.strings_to_categoricals.html>`__) 

    Only affects string annotations that lead to less categories than the total number of observations.

- Transpose the data matrix, leaving observations and variables interchanged (`transpose method <https://anndata.readthedocs.io/en/latest/anndata.AnnData.transpose.html>`__)

    Data matrix is transposed, observations and variables are interchanged.

- Add annotation for variables or observations 

- Filter data variables or observations, by index or key

- Freeze the current state into the 'raw' attribute

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