annotate decoupler_pathway_inference.xml @ 10:97c2c52a7ab4 draft default tip

planemo upload for repository https://github.com/ebi-gene-expression-group/container-galaxy-sc-tertiary/ commit b581a5b4ba88c5bf06f6223ba9aec51a8564796c

<tool id="decoupler_pathway_inference" name="Decoupler Pathway Inference" version="1.4.0+galaxy3" profile="20.05" license="MIT">
    <description>
        of functional genesets/pathways for scRNA-seq data.
    </description>
    <requirements>
        <requirement type="package" version="1.4.0">decoupler</requirement>
    </requirements>
    <command>
        #if $inp.format == 'h5ad':
            #set $input_fname = "input.h5ad"
        #else:
            #set $input_fname = "input.tsv"
        #end if
        ln -s '$input' '$input_fname';

        python '$__tool_directory__/decoupler_pathway_inference.py'
            -i '$input_fname'
            -n '$input_network_file'
            --min_n "$min_n"
            --method '$method'
            
            --source '$source'
            --target '$target'
            --weight '$weight'
            #if str($inp.format) == "tabular":
                #if $inp.stat_field:
                    --stat "${inp.stat_field}"
                #end if
                #if $inp.p_value_column:
                    --p_value_column "${inp.p_value_column}"
                    --p_value_threshold "${inp.p_value_threshold}"
                #end if
            #else:
                #if $inp.gene_symbols_field:
                    --var_gene_symbols_field "${inp.gene_symbols_field}"
                #end if
                #if $inp.use_raw:
                    ${inp.use_raw}
                #end if
                #if $inp.write_activities_path:
                    ${inp.write_activities_path}
                #end if
            #end if
            --output "inference"
            
    </command>
    <inputs>
        <param name="input" type="data" format="h5ad,tabular" label="Input AnnData/Expression file"/>
        <param name="input_network_file" type="data" format="tabular" label="Input Network file" help="Tabular file with columns Source, Target and Weight. A source gene/pathway regulates/contains a target gene, weights can be either positive or negative. The source element needs to be part of the network, the target is a gene in the network and in the dataset"/>
        <param name="min_n" type="integer" min="0" value="5" label="Minimum targets per source." help="If targets are less than minimum, sources are removed"/>
        <conditional name="inp">
            <param name="format" type="select" label="Input Format" help="Whether the provided file is AnnData or a Table of differential expression results (usually from bulk).">
                <option value="h5ad">AnnData</option>
                <option value="tabular">Differential Expression Table</option>
            </param>
            <when value="h5ad">
                <param name="use_raw" type="boolean" truevalue="--use_raw" falsevalue="" checked="false" label="Use the raw part of the AnnData object"/>
                <param name="write_activities_path" type="boolean" truevalue="--activities_path anndata_activities_path.h5ad" falsevalue="" checked="true" label="Write the activities AnnData object." help="Contains the MLM/ULM/Consensus activity results for each pathway and each cell in the main matrix, it is not a replacement of the original AnnData provided as input."/>
                <param name="gene_symbols_field" type="text" optional="true" label="Gene symbols field" help="The field in the AnnData var table where gene symbols are stored."/>
            </when>
            <when value="tabular">
                <param name="stat_field" type="text" label="Statistic column name" optional="false" help="Defines which column will be passed to the decoupler method, usually you want something like the log2FC or the t-stat (this must be a column in your table)"/>
                <param argument="--p_value_column" type="text" label="P-value/FDR column name" help="Defines which column will be passed to the decoupler method as p-value, usually you want something like the log2FC or the t-stat (this must be a column in your table)"/>
                <param argument="--p_value_threshold" value="0.05" type="float" label="P-value/FDR thresholds" help="Will filter out any rows in the file that are above the value (in the set P-value/FDR column)"/>
            </when>
        </conditional>
        <param name="method" type="select" label="Activity inference method">
            <option value="mlm" selected="true">Multivariate linear model (MLM)</option>
            <option value="ulm">Univariate linear model (ULM)</option>
            <option value="consensus">Consensus (use for TFs with CollecTri)</option>
        </param>
        <param name="source" type="text" value="source" label="Column name in network with source nodes." help="Usually the regulators. If empty then default is 'source' is used."/>
        <param name="target" type="text" value="target" label="Column name in network with target nodes." help="Usually the regulated genes. If empty then default is 'target' is used."/>
        <param name="weight" type="text" value="weight" label="Column name in network with weight." help="If empty then default is 'weight' is used."/>
    </inputs>
    <outputs>
        <data name="output_ad" format="h5ad" from_work_dir="anndata_activities_path.h5ad" label="${tool.name} on ${on_string}: Regulators/Pathways activity AnnData file">
            <filter>inp['format'] == "h5ad" and inp['write_activities_path'] is True</filter>
        </data>
        <data name="output_table" format="tabular" from_work_dir="inference.tsv" label="${tool.name} on ${on_string}: Output estimate table"/>
    </outputs>
    <tests>
        <!-- Hint: You can use [ctrl+alt+t] after defining the inputs/outputs to auto-scaffold some basic test cases. -->
        <test expect_num_outputs="2">
            <param name="input" value="pbmc3k_processed.h5ad"/>
            <param name="inp|format" value="h5ad"/>
            <param name="input_network_file" value="progeny_test.tsv"/>
            <param name="min_n" value="0"/>
            <param name="method" value="mlm"/>
            <param name="inp|use_raw" value="false"/>
            <param name="inp|write_activities_path" value="true"/>
            <param name="source" value="source"/>
            <param name="target" value="target"/>
            <param name="weight" value="weight"/>
            <output name="output_ad">
                <assert_contents>
                    <has_h5_keys keys="obsm/mlm_estimate"/>
                </assert_contents>
            </output>
            <output name="output_table">
                <assert_contents>
                    <has_n_columns n="5"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="2">
            <param name="input" value="pbmc3k_processed.h5ad"/>
            <param name="inp|format" value="h5ad"/>
            <param name="input_network_file" value="progeny_test_2.tsv"/>
            <param name="min_n" value="0"/>
            <param name="method" value="ulm"/>
            <param name="inp|use_raw" value="false"/>
            <param name="inp|write_activities_path" value="true"/>
            <param name="source" value="source"/>
            <param name="target" value="target"/>
            <param name="weight" value="weight"/>
            <output name="output_ad">
                <assert_contents>
                    <has_h5_keys keys="obsm/ulm_estimate"/>
                </assert_contents>
            </output>
            <output name="output_table">
                <assert_contents>
                    <has_n_columns n="5"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="2">
            <param name="input" value="mito_counted_anndata.h5ad"/>
            <param name="inp|format" value="h5ad"/>
            <param name="input_network_file" value="mouse_progeny.tsv"/>
            <param name="min_n" value="0"/>
            <param name="method" value="ulm"/>
            <param name="inp|use_raw" value="false"/>
            <param name="inp|write_activities_path" value="true"/>
            <param name="source" value="source"/>
            <param name="target" value="target"/>
            <param name="weight" value="weight"/>
            <param name="inp|gene_symbols_field" value="Symbol"/>
            <output name="output_ad">
                <assert_contents>
                    <has_h5_keys keys="obsm/ulm_estimate"/>
                </assert_contents>
            </output>
            <output name="output_table">
                <assert_contents>
                    <has_n_columns n="29"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="1">
            <param name="input" value="diff_exp_result.tab"/>
            <param name="inp|format" value="tabular"/>
            <param name="input_network_file" value="progeny_test.tsv"/>
            <param name="min_n" value="0"/>
            <param name="method" value="mlm"/>
            <param name="inp|stat_field" value="log2FoldChange"/>
            <param name="inp|write_activities_path" value="false"/>
            <param name="source" value="source"/>
            <param name="target" value="target"/>
            <param name="weight" value="weight"/>
            <output name="output_table">
                <assert_contents>
                    <has_n_columns n="3"/>
                </assert_contents>
            </output>
        </test>
    </tests>
    <help>
**What it does**

Usage
.....


**Description**

This tool extracts pathway activity inference using decoupler. For more information on the underlying algorithms, 
the `decoupler documentation`_ , in particular the Pathway Activity
and Transcription factor activity inference sections.

.. _`decoupler documentation`: https://decoupler-py.readthedocs.io/en/latest/

**Input**  

The tool accepts two types of input files:

1. An AnnData object in H5AD format:

    - The H5AD file can contain raw or normalized data.
    - You can specify whether to use the raw data in the AnnData object instead of the X matrix using the "use_raw" parameter.
    - Minimum of targets per source can be specified using "min_n".

2. A tabular file with differential expression data:

    - The file should have genes in rows (and the first column by the gene symbols).
    - The file needs a header, that is, columns names for every column.
    - Columns must include at least fields similar to log2FC and a p-value or FDR field.
    - If this file is provided, the tool will score each source in the network file according to the differential expression of the provided genes.

The tool also requires a network file containing a collection of pathways and their target genes, with weights for each interaction.

Example of a network file:

+---------+--------+--------+
| source  | target | weight |
+=========+========+========+
| T1      | G01    | 1.0    |
+---------+--------+--------+
| T1      | G02    | 1.0    |
+---------+--------+--------+
| T1      | G03    | 0.7    |
+---------+--------+--------+
| T2      | G04    | 1.0    |
+---------+--------+--------+
| T2      | G06    | -0.5   |
+---------+--------+--------+


**Output**

Depending on the input file type, the tool outputs:

- If an AnnData file is used:

    - An AnnData object containing the scores in the "obs" field.
    - Tab-separated text files containing the scores for each cell.
    - If the "write_activities_path" parameter is set to "true", the tool will write the modified AnnData object to an H5AD file.  
    - If the "write_inference" parameter is set to "true", the tool will output a tab-separated text file containing the scores for each cell.

- If a tabular differential expression file is used:

    - A tab-separated text file where each source in the network file is scored according to the differential expression of the provided genes.

    </help>
    <citations>
        <citation type="doi">10.1093/bioadv/vbac016 </citation>
    </citations>
</tool>