annotate filter.xml @ 0:6ea5a05a260a draft

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/scanpy/ commit 92f85afaed0097d1879317a9f513093fce5481d6

<tool id="scanpy_filter" name="Filter with scanpy" version="@galaxy_version@">
    <description></description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements"/>
    <expand macro="version_command"/>
    <command detect_errors="exit_code"><![CDATA[
@CMD@
      ]]></command>
    <configfiles>
        <configfile name="script_file"><![CDATA[
@CMD_imports@
@CMD_read_inputs@

#if $method.method == 'pp.filter_cells'
res = sc.pp.filter_cells(
    #if $modify_anndata.modify_anndata == 'true'
    adata,
    #else
    adata.X,
    #end if
    #if $method.filter.filter == 'min_counts'
    min_counts=$method.filter.min_counts,
    #elif $method.filter.filter == 'max_counts'
    max_counts=$method.filter.max_counts,
    #elif $method.filter.filter == 'min_genes'
    min_genes=$method.filter.min_genes,
    #elif $method.filter.filter == 'max_genes'
    max_genes=$method.filter.max_genes,
    #end if
    copy=False)

    #if $modify_anndata.modify_anndata == 'true'
df = adata.obs
    #else
df = pd.DataFrame(data=dict(cell_subset=res[0], number_per_cell=res[1]))
    #end if

    #if $method.filter.filter == 'min_counts' or $method.filter.filter == 'max_counts'
df.to_csv('$counts_per_cell', sep='\t')
    #elif $method.filter.filter == 'min_genes' or $method.filter.filter == 'max_genes'
df.to_csv('$genes_per_cell', sep='\t')
    #end if

#elif $method.method == 'pp.filter_genes'
res = sc.pp.filter_genes(
    #if $modify_anndata.modify_anndata == 'true'
    adata,
    #else
    adata.X,
    #end if
    #if $method.filter.filter == 'min_counts'
    min_counts=$method.filter.min_counts,
    #elif $method.filter.filter == 'max_counts'
    max_counts=$method.filter.max_counts,
    #elif $method.filter.filter == 'min_cells'
    min_cells=$method.filter.min_cells,
    #elif $method.filter.filter == 'max_cells'
    max_cells=$method.filter.max_cells,
    #end if
    copy=False)

    #if $modify_anndata.modify_anndata == 'true'
df = adata.var
    #else
df = pd.DataFrame(data=dict(gene_subset=res[0], number_per_gene=res[1]))
    #end if

    #if $method.filter.filter == 'min_counts' or $method.filter.filter == 'max_counts'
df.to_csv('$counts_per_gene', sep='\t')
    #elif $method.filter.filter == 'min_cells' or $method.filter.filter == 'max_cells'
df.to_csv('$cells_per_gene', sep='\t')
    #end if

#elif $method.method == 'pp.filter_genes_dispersion'
res = sc.pp.filter_genes_dispersion(
    #if $modify_anndata.modify_anndata == 'true'
    adata,
    #else
    adata.X,
    #end if
    flavor='$method.flavor.flavor',
    #if $method.flavor.flavor=='seurat'
    min_mean=$method.flavor.min_mean,
    max_mean=$method.flavor.max_mean,
    min_disp=$method.flavor.min_disp,
        #if $method.flavor.max_disp
    max_disp=$method.flavor.max_disp,
        #end if
    #else
    n_top_genes=$method.flavor.n_top_genes,
    #end if
    n_bins=$method.n_bins,
    log=$method.log,
    copy=False)

    #if $modify_anndata.modify_anndata == 'true'
adata.var.to_csv('$per_gene', sep='\t')
    #else
pd.DataFrame(res).to_csv('$per_gene', sep='\t')
    #end if

#elif $method.method == 'pp.subsample'
sc.pp.subsample(
    data=adata,
    #if $method.type.type == 'fraction'
    fraction=$method.type.fraction,
    #else if $method.type.type == 'n_obs'
    n_obs=$method.type.n_obs,
    #end if
    random_state=$method.random_state,
    copy=False)

#end if

@CMD_anndata_write_modify_outputs@
]]></configfile>
    </configfiles>
    <inputs>
        <expand macro="inputs_anndata"/>
        <conditional name="method">
            <param argument="method" type="select" label="Method used for filtering">
                <option value="pp.filter_cells">Filter cell outliers based on counts and numbers of genes expressed, using `pp.filter_cells`</option>
                <option value="pp.filter_genes">Filter genes based on number of cells or counts, using `pp.filter_genes`</option>
                <option value="pp.filter_genes_dispersion">Extract highly variable genes, using `pp.filter_genes_dispersion`</option>
                <!--<option value="pp.highly_variable_genes">, using `tl.highly_variable_genes`</option>!-->
                <option value="pp.subsample">Subsample to a fraction of the number of observations, using `pp.subsample`</option>
                <!--<option value="queries.gene_coordinates">, using `queries.gene_coordinates`</option>!-->
                <!--<option value="queries.mitochondrial_genes">, using `queries.mitochondrial_genes`</option>!-->
            </param>
            <when value="pp.filter_cells">
                <conditional name="filter">
                    <param argument="filter" type="select" label="Filter">
                        <option value="min_counts">Minimum number of counts</option>
                        <option value="max_counts">Maximum number of counts</option>
                        <option value="min_genes">Minimum number of genes expressed</option>
                        <option value="max_genes">Maximum number of genes expressed</option>
                    </param>
                    <when value="min_counts">
                        <param argument="min_counts" type="integer" min="0" value="" label="Minimum number of counts required for a cell to pass filtering" help=""/>
                    </when>
                    <when value="max_counts">
                        <param argument="max_counts" type="integer" min="0" value="" label="Maximum number of counts required for a cell to pass filtering" help=""/>
                    </when>
                    <when value="min_genes">
                        <param argument="min_genes" type="integer" min="0" value="" label="Minimum number of genes expressed required for a cell to pass filtering" help=""/>
                    </when>    
                    <when value="max_genes">
                        <param argument="max_genes" type="integer" min="0" value="" label="Maximum number of genes expressed required for a cell to pass filtering" help=""/>
                    </when>
                </conditional>
            </when>
            <when value="pp.filter_genes">
                <conditional name="filter">
                    <param argument="filter" type="select" label="Filter">
                        <option value="min_counts">Minimum number of counts</option>
                        <option value="max_counts">Maximum number of counts</option>
                        <option value="min_cells">Minimum number of cells expressed</option>
                        <option value="max_cells">Maximum number of cells expressed</option>
                    </param>
                    <when value="min_counts">
                        <param argument="min_counts" type="integer" min="0" value="" label="Minimum number of counts required for a gene to pass filtering" help=""/>
                    </when>
                    <when value="max_counts">
                        <param argument="max_counts" type="integer" min="0" value="" label="Maximum number of counts required for a gene to pass filtering" help=""/>
                    </when>
                    <when value="min_cells">
                        <param argument="min_cells" type="integer" min="0" value="" label="Minimum number of cells expressed required for a gene to pass filtering" help=""/>
                    </when>    
                    <when value="max_cells">
                        <param argument="max_cells" type="integer" min="0" value="" label="Maximum number of cells expressed required for a gene to pass filtering" help=""/>
                    </when>
                </conditional>
            </when>
            <when value="pp.filter_genes_dispersion">
                <conditional name='flavor'>
                    <param argument="flavor" type="select" label="Flavor for computing normalized dispersion" help="">
                        <option value="seurat">seurat: expects non-logarithmized data</option>
                        <option value="cell_ranger">cell_ranger: usually called for logarithmized data</option>
                    </param>
                    <when value="seurat">
                        <param argument="min_mean" type="float" value="0.0125" label="Minimal mean cutoff" help=""/>
                        <param argument="max_mean" type="float" value="3" label="Maximal mean cutoff" help=""/>
                        <param argument="min_disp" type="float" value="0.5" label="Minimal normalized dispersion cutoff" help=""/>
                        <param argument="max_disp" type="float" value="" optional="true" label="Maximal normalized dispersion cutoff" help=""/>
                    </when>
                    <when value="cell_ranger">
                        <param argument="n_top_genes" type="integer" value="" label="Number of highly-variable genes to keep" help=""/>
                    </when>
                </conditional>
                <param argument="n_bins" type="integer" value="20" label="Number of bins for binning the mean gene expression" help="Normalization is done with respect to each bin. If just a single gene falls into a bin, the normalized dispersion is artificially set to 1"/>
                <expand macro="param_log"/>
            </when>
            <when value="pp.subsample">
                <conditional name="type">
                    <param name="type" type="select" label="Type of subsampling">
                        <option value="fraction">By fraction</option>
                        <option value="n_obs">By number of observation</option>
                    </param>
                    <when value="fraction">
                        <param argument="fraction" type="float" value="" label="Subsample to this `fraction` of the number of observations" help=""/>
                    </when>
                    <when value="n_obs">
                        <param argument="n_obs" type="integer" min="0" value="" label="Subsample to this number of observations" help=""/>
                    </when>
                </conditional>
                <param argument="random_state" type="integer" value="0" label="Random seed to change subsampling" help=""/>
            </when>
        </conditional>
        <expand macro="anndata_modify_output_input"/>
    </inputs>
    <outputs>
        <expand macro="anndata_modify_outputs"/>
        <!-- for pp.filter_cells -->
        <data name="counts_per_cell" format="tabular" label="${tool.name} on ${on_string}: Counts per cells after filtering">
            <filter>method['method'] == 'pp.filter_cells' and (method['filter']['filter'] == 'min_counts' or method['filter']['filter'] == 'max_counts')</filter>
        </data>
        <data name="genes_per_cell" format="tabular" label="${tool.name} on ${on_string}: Number of genes per cell after filtering">
            <filter>method['method'] == 'pp.filter_cells' and (method['filter']['filter'] == 'min_genes' or method['filter']['filter'] == 'max_genes')</filter>
        </data>
        <!-- for pp.filter_genes -->
        <data name="counts_per_gene" format="tabular" label="${tool.name} on ${on_string}: Counts per genes after filtering">
            <filter>method['method'] == 'pp.filter_genes' and (method['filter']['filter'] == 'min_counts' or method['filter']['filter'] == 'max_counts')</filter>
        </data>
        <data name="cells_per_gene" format="tabular" label="${tool.name} on ${on_string}: Number of cells per genes after filtering">
            <filter>method['method'] == 'pp.filter_genes' and (method['filter']['filter'] == 'min_cells' or method['filter']['filter'] == 'max_cells')</filter>
        </data>
        <!-- for pp.filter_genes_dispersion -->
        <data name="per_gene" format="tabular" label="${tool.name} on ${on_string}: Means, dispersions and normalized dispersions per gene">
            <filter>method['method'] == 'pp.filter_genes_dispersion'</filter>
        </data>
    </outputs>
    <tests>
        <test expect_num_outputs="2">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="krumsiek11.h5ad" />
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.filter_cells"/>
                <conditional name="filter">
                    <param name="filter" value="min_counts"/>
                    <param name="min_counts" value="3"/>
                </conditional>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="true"/>
                <param name="anndata_output_format" value="h5ad" />
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.filter_cells"/>
                <has_text_matching expression="min_counts=3"/>
            </assert_stdout>
            <output name="anndata_out_h5ad" file="pp.filter_cells.krumsiek11-min_counts.h5ad" ftype="h5" compare="sim_size"/>
            <output name="counts_per_cell">
                <assert_contents>
                    <has_text_matching expression="cell_type\tn_counts" />
                    <has_text_matching expression="46\tprogenitor\t3.028" />
                    <has_text_matching expression="85\tEry\t3.7001" />
                    <has_text_matching expression="150\tMk\t4.095" />
                    <has_n_columns n="3" />
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="2">
            <conditional name="input">
                <param name="format" value="loom" />
                <param name="adata" value="krumsiek11.loom" />
                <param name="sparse" value="True"/>
                <param name="cleanup" value="False"/>
                <param name="x_name"  value="spliced"/>
                <param name="obs_names" value="CellID" />
                <param name="var_names" value="Gene"/>
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.filter_cells"/>
                <conditional name="filter">
                    <param name="filter" value="min_counts"/>
                    <param name="min_counts" value="3"/>
                </conditional>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="true"/>
                <param name="anndata_output_format" value="loom" />
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.filter_cells"/>
                <has_text_matching expression="min_counts=3"/>
            </assert_stdout>
            <output name="anndata_out_loom" file="pp.filter_cells.krumsiek11-min_counts.loom" ftype="loom" compare="sim_size"/>
            <output name="counts_per_cell">
                <assert_contents>
                    <has_text_matching expression="cell_type\tn_counts" />
                    <has_text_matching expression="46\tprogenitor\t3.028" />
                    <has_text_matching expression="85\tEry\t3.7001" />
                    <has_text_matching expression="97\tMo\t3.925" />
                    <has_text_matching expression="150\tMk\t4.095" />
                    <has_n_columns n="3" />
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="1">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="krumsiek11.h5ad"/>
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.filter_cells"/>
                <conditional name="filter">
                    <param name="filter" value="max_genes"/>
                    <param name="max_genes" value="100"/>
                </conditional>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="false"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.filter_cells"/>
                <has_text_matching expression="adata.X"/>
                <has_text_matching expression="max_genes=100"/>
            </assert_stdout>
            <output name="genes_per_cell" file="pp.filter_cells.number_per_cell.krumsiek11-max_genes.tabular"/>
        </test>
        <test expect_num_outputs="2">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="krumsiek11.h5ad" />
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.filter_genes"/>
                <conditional name="filter">
                    <param name="filter" value="min_counts"/>
                    <param name="min_counts" value="3"/>
                </conditional>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="true"/>
                <param name="anndata_output_format" value="h5ad" />
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.filter_genes"/>
                <has_text_matching expression="min_counts=3"/>
            </assert_stdout>
            <output name="anndata_out_h5ad" file="pp.filter_genes.krumsiek11-min_counts.h5ad" ftype="h5" compare="sim_size"/>
            <output name="counts_per_gene" file="pp.filter_genes.number_per_gene.krumsiek11-min_counts.tabular"/>
        </test>
        <test expect_num_outputs="1">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="pbmc68k_reduced.h5ad"/>
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.filter_genes"/>
                <conditional name="filter">
                    <param name="filter" value="max_cells"/>
                    <param name="max_cells" value="500"/>
                </conditional>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="false"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.filter_genes"/>
                <has_text_matching expression="adata.X"/>
                <has_text_matching expression="max_cells=500"/>
            </assert_stdout>
            <output name="cells_per_gene" file="pp.filter_genes.number_per_gene.pbmc68k_reduced-max_cells.tabular"/>
        </test>
        <test expect_num_outputs="2">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="krumsiek11.h5ad" />
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.filter_genes_dispersion"/>
                <conditional name="flavor">
                    <param name="flavor" value="seurat"/>
                    <param name="min_mean" value="0.0125"/>
                    <param name="max_mean" value="3"/>
                    <param name="min_disp" value="0.5"/>
                </conditional>
                <param name="n_bins" value="20" />
                <param name="log" value="true"/>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="true"/>
                <param name="anndata_output_format" value="h5ad" />
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.filter_genes_dispersion"/>
                <has_text_matching expression="flavor='seurat'"/>
                <has_text_matching expression="min_mean=0.0125"/>
                <has_text_matching expression="max_mean=3.0"/>
                <has_text_matching expression="min_disp=0.5"/>
                <has_text_matching expression="n_bins=20"/>
                <has_text_matching expression="log=True"/>
            </assert_stdout>
            <output name="anndata_out_h5ad" file="pp.filter_genes_dispersion.krumsiek11-seurat.h5ad" ftype="h5" compare="sim_size"/>
            <output name="per_gene" file="pp.filter_genes_dispersion.per_gene.krumsiek11-seurat.tabular"/>
        </test>
        <test expect_num_outputs="1">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="krumsiek11.h5ad" />
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.filter_genes_dispersion"/>
                <conditional name="flavor">
                    <param name="flavor" value="cell_ranger"/>
                    <param name="n_top_genes" value="2"/>
                </conditional>
                <param name="n_bins" value="20"/>
                <param name="log" value="true"/>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="false"/>
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.filter_genes_dispersion"/>
                <has_text_matching expression="flavor='cell_ranger'"/>
                <has_text_matching expression="n_top_genes=2"/>
                <has_text_matching expression="n_bins=20"/>
                <has_text_matching expression="og=True"/>
            </assert_stdout>
            <output name="per_gene" file="pp.filter_genes_dispersion.per_gene.krumsiek11-cell_ranger.tabular"/>
        </test>
        <test expect_num_outputs="1">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="krumsiek11.h5ad" />
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.subsample"/>
                <conditional name="type">
                    <param name="type" value="fraction" />
                    <param name="fraction" value="0.5"/>
                </conditional>
                <param name="random_state" value="0"/>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="true"/>
                <param name="anndata_output_format" value="h5ad" />
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.subsample"/>
                <has_text_matching expression="fraction=0.5"/>
                <has_text_matching expression="random_state=0"/>
            </assert_stdout>
            <output name="anndata_out_h5ad" file="pp.subsample.krumsiek11_fraction.h5ad" ftype="h5" compare="sim_size"/>
        </test>
        <test expect_num_outputs="1">
            <conditional name="input">
                <param name="format" value="h5ad" />
                <param name="adata" value="krumsiek11.h5ad" />
            </conditional>
            <conditional name="method">
                <param name="method" value="pp.subsample"/>
                <conditional name="type">
                    <param name="type" value="n_obs" />
                    <param name="n_obs" value="10"/>
                </conditional>
                <param name="random_state" value="0"/>
            </conditional>
            <conditional name="modify_anndata">
                <param name="modify_anndata" value="true"/>
                <param name="anndata_output_format" value="h5ad" />
            </conditional>
            <assert_stdout>
                <has_text_matching expression="sc.pp.subsample"/>
                <has_text_matching expression="n_obs=10"/>
                <has_text_matching expression="random_state=0"/>
            </assert_stdout>
            <output name="anndata_out_h5ad" file="pp.subsample.krumsiek11_n_obs.h5ad" ftype="h5" compare="sim_size"/>
        </test>
    </tests>
    <help><![CDATA[

Filter cells outliers based on counts and numbers of genes expressed (`pp.filter_cells`)
========================================================================================

For instance, only keep cells with at least `min_counts` counts or
`min_genes` genes expressed. This is to filter measurement outliers, i.e.,
"unreliable" observations.

Only provide one of the optional parameters `min_counts`, `min_genes`,
`max_counts`, `max_genes` per call.

More details on the `scanpy documentation
<https://scanpy.readthedocs.io/en/latest/api/scanpy.api.pp.filter_cells.html#scanpy.api.pp.filter_cells>`__

Return
------

number_per_cell : Number per cell (either `n_counts` or `n_genes` per cell)


Filter genes based on number of cells or counts (`pp.filter_genes`)
===================================================================

Keep genes that have at least `min_counts` counts or are expressed in at
least `min_cells` cells or have at most `max_counts` counts or are expressed
in at most `max_cells` cells.

Only provide one of the optional parameters `min_counts`, `min_cells`,
`max_counts`, `max_cells` per call.

More details on the `scanpy documentation
<https://scanpy.readthedocs.io/en/latest/api/scanpy.api.pp.filter_genes.html#scanpy.api.pp.filter_genes>`__

Return
------

number_per_gene : Number per genes (either `n_counts` or `n_genes` per cell)


Extract highly variable genes (`pp.filter_genes_dispersion`)
============================================================

If trying out parameters, pass the data matrix instead of AnnData.

Depending on `flavor`, this reproduces the R-implementations of Seurat and Cell Ranger.

The normalized dispersion is obtained by scaling with the mean and standard
deviation of the dispersions for genes falling into a given bin for mean
expression of genes. This means that for each bin of mean expression, highly
variable genes are selected.

Use `flavor='cell_ranger'` with care and in the same way as in `pp.recipe_zheng17`.

More details on the `scanpy documentation
<https://scanpy.readthedocs.io/en/latest/api/scanpy.api.pp.filter_genes_dispersion.html#scanpy.api.pp.filter_genes_dispersion>`__

Returns
-------
- The annotated matrix filtered, with the annotations
- A table with the means, dispersions, and normalized dispersions per gene, logarithmized when `log` is `True`.


Subsample to a fraction of the number of observations (`pp.subsample`)
======================================================================

More details on the `scanpy documentation
<https://scanpy.readthedocs.io/en/latest/api/scanpy.api.pp.subsample.html#scanpy.api.pp.subsample>`__


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