comparison abims_correlation_analysis.xml @ 0:58997c28b268 draft default tip

"planemo upload for repository https://github.com/workflow4metabolomics/tools-metabolomics/blob/master/tools/correlation_analysis/ commit 35a01e4ef59a91f43d0b1de1d08db29dcc7aae1e"

0:58997c28b268

<tool id="correlation_analysis" name="Metabolites Correlation Analysis" version="1.0.1+galaxy0" >

    <description>to highlight ion correlations considering PC-groups</description>

    <requirements>
        <requirement type="package" version="1.1_5">r-batch</requirement>
        <requirement type="package" version="0.8.8">r-reshape</requirement>
        <requirement type="package" version="7.3_53">r-mass</requirement>
    </requirements>

    <command detect_errors='exit_code'>
        Rscript '$__tool_directory__/correlation_analysis.r'

        #if $cond_input_type.select_input_type == "select_input_from_w4m" and $cond_input_type.cond_function.select_funtion == "sort_only" :
            sorting 1 variable_metadata '$cond_input_type.variableMetadata'
            data_matrix '$cond_input_type.dataMatrix'
            sample_metadata '$cond_input_type.sampleMetadata'
            corrdel 0
            param_correlation ""
            param_cytoscape ""
            matrix_corr 0
            user_matrix_corr ""
            corr_method ""
        #end if
        #if $cond_input_type.select_input_type == "select_input_from_w4m" and $cond_input_type.cond_function.select_funtion == "sort_and_corr" :
            sorting 1
            variable_metadata '$cond_input_type.variableMetadata'
            data_matrix '$cond_input_type.dataMatrix'
            sample_metadata '$cond_input_type.sampleMetadata'
            corrdel 1
            param_correlation $cond_input_type.cond_function.param_correlation
            param_cytoscape $cond_input_type.cond_function.param_cytoscape
            matrix_corr 0
            user_matrix_corr ""
            corr_method $cond_input_type.cond_function.corr_method
        #end if
        ##Create correlation matrix from a user table file.##
        #if $cond_input_type.select_input_type == "select_input_other" :
            sorting 0
            variable_metadata ""
            data_matrix ""
            sample_metadata ""
            corrdel 0
            param_correlation ""
            param_cytoscape $cond_input_type.param_cytoscape
            matrix_corr 1
            user_matrix_corr '$cond_input_type.user_matrix_corr'
            corr_method $cond_input_type.corr_method
        #end if

    </command>

    <inputs>


         <conditional name="cond_input_type" >
            <param name="select_input_type" type="select" label="Choice of your input files" help="" >
                <option value="select_input_from_w4m" selected="true">Files from the metabolomic workflow</option>
                <option value="select_input_other" >Your table file</option>
            </param>
            <when value="select_input_from_w4m">
                <param name="dataMatrix" type="data" label="Data matrix" format="tabular" help="dataMatrix file from the CAMERA.annotate step for example" />
                <param name="sampleMetadata" type="data" label="Sample metadata" format="tabular" help="sampleMetadata file from the xcms.xcmsSet step for example" />
                <param name="variableMetadata" type="data" label="Variable metadata" format="tabular" help="variableMetadata file from the CAMERA.annotate step for example" />
                <conditional name="cond_function" >
                    <param name="select_funtion" type="select" label="Function to be used" help="" >
                        <option value="sort_only" selected="true">Sorting your table</option>
                        <option value="sort_and_corr" >Sorting your table and doing correlation analysis</option>
                    </param>
                    <when value="sort_only" />
                    <when value="sort_and_corr">
                         <param name="corrdel" type="hidden" value="1"/>
                         <param name="param_correlation" type="float" label="Correlation threshold for pcgroup" value="0.60" help="Threshold value for selecting edges (i.e. correlations) that will be exported to the Cytoscape sif format file" />
                         <param name="corr_method" type="select" label="Choice of the correlation method" help="" >
                            <option value="pearson" selected="true">pearson</option>
                            <option value="kendall">kendall</option>
                            <option value="spearman">spearman</option>
                         </param>
                         <param name="param_cytoscape" type="float" label="Cytoscape correlation threshold" value="0.75" help="Choose a threshold value for selecting metabolites that will be exported to a cytoscape sif format" />
                    </when>
                </conditional>
            </when>

            <when value="select_input_other">
                <param name="user_matrix_corr" type="data" label="Your table file (tabular format)" format="tabular" help="Your metabolites (variables) intensity table file (tabular format)" />
                <param name="corr_method" type="select" label="Choice of the correlation method" help="" >
                    <option value="pearson" selected="true">pearson</option>
                    <option value="kendall">kendall</option>
                    <option value="spearman">spearman</option>
                </param>
                <param name="param_cytoscape" type="float" label="Cytoscape correlation threshold" value="0.75" help="Threshold value for selecting edges (i.e. correlations) that will be exported to the Cytoscape sif format file" />
            </when>
        </conditional>
    </inputs>

    <outputs>
        <data name="sorted_table" format="tabular" from_work_dir="sorted_table.tsv" label="sorted_variableMetadata.tsv">
             <filter>(cond_input_type['select_input_type'] == 'select_input_from_w4m')</filter>
        </data>
        <data name="correlation_matrix_selected" format="tabular" from_work_dir="correlation_matrix_selected.tsv" label="correlation_matrix_selected.tsv">
            <filter>(cond_input_type['select_input_type'] == 'select_input_from_w4m' and cond_input_type['cond_function']['select_funtion']== 'sort_and_corr' and cond_input_type['cond_function']['corrdel']== '1' )</filter>
        </data>
        <data name="siff_table" format="tabular" from_work_dir="siff_table.tsv" label="sif_table.tsv">
            <filter>(cond_input_type['select_input_type'] == 'select_input_from_w4m' and cond_input_type['cond_function']['select_funtion']== 'sort_and_corr' and cond_input_type['cond_function']['corrdel']== '1' )</filter>
        </data>
        <data name="correlation_matrix_user" format="tabular" from_work_dir="correlation_matrix.tsv" label="correlation_matrix.tsv">
            <filter>(cond_input_type['select_input_type'] == 'select_input_other')</filter>
        </data>
        <data name="siff_table_user_user" format="tabular" from_work_dir="siff_table.tsv" label="sif_table.tsv">
            <filter>(cond_input_type['select_input_type'] == 'select_input_other')</filter>
        </data>
    </outputs>

    <tests>
        <test expect_num_outputs="3">
            <conditional name="cond_input_type" >
                <param name="select_input_type" value="select_input_from_w4m" />
                <param name="dataMatrix" value="in_DM1.tabular" ftype="tabular" />
                <param name="sampleMetadata" value="in_SM1.tabular" ftype="tabular" />
                <param name="variableMetadata" value="in_VM1.tabular" ftype="tabular" />
                <conditional name="cond_function" >
                    <param name="select_funtion" value="sort_and_corr" />
                </conditional>
            </conditional>
            <output name="sorted_table" file="out_VM1.tabular"/>
            <output name="correlation_matrix_selected" file="out_corr1.tabular"/>
            <output name="siff_table" file="out_sif1.tabular"/>
        </test>
    </tests>

    <help>

.. class:: infomark

**Authors** Antoine Gravot (Protocole conception) and Misharl Monsoor (for initial galaxy wrapper and R script).

**Additional W4M contributors** ABiMS TEAM (SU/CNRS - Station biologique de Roscoff) and PFEM (INRAE - MetaboHUB)

---------------------------------------------------

================================
Metabolites correlation analysis
================================

-----------
Description
-----------

This tool takes as inputs either tabular table files from the metabolomic workflow (variableMetadata, dataMatrix and sampleMetadata) or a table file of your own
and can execute three different functions ("sorting", "corrdel" and "corr_matrix").

**The "sorting" function:** *used for metabolomic workflow*

 | 1) First of all, it sorts the data by pcgroup.
 | 2) It computes the mean operation of all the signal values of the metabolites by sample, and put the results in a new column "signal_moy".
 | 3) It finally creates a tabular output "sorted_variableMetadata.tsv".

**The "corrdel" function:** *used for metabolomic workflow*

 | **For each pcgroup** of the previous sorted tabular file "sorted_table.tsv", it does the following things:
 | - it computes a correlation matrix
 | - it determines the metabolites which are not correlated to others from the same pcgroup based on the threshold value filled in the "Correlation threshold for pcgroup" parameter
 | - the metabolites are sorted by the mean signal intensity (form the highest to the lowest), and each metabolite is tested to the previous ones in the list ; if the tested metabolite is at least correlated to one previous one, it is tagged as DEL (for "deleted", written in a column called "suppress")
 |
 | It creates two additional tabular files:
 | - "correlation_matrix_selected.tsv" (correlation matrix of selected metabolites only)
 | - "sif_table.tsv" (for visualization in CytoScape, based on selected metabolites and "Cytoscape correlation threshold" filled value)


**The "corr_matrix" function:** *used for user table file*

 | It computes a correlation matrix named "correlation_matrix.tsv" and creates a sif file named "sif_table.tsv" (for visualization in CytoScape).



-----------------
Workflow position
-----------------


**Examples of upstream tools**

+---------------------------+--------------------------+--------+------------------------+
| Name                      | Output file              | Format | parameter              |
+===========================+==========================+========+========================+
|xcms findChromPeaks Merger |sampleMetada.tsv          | Tabular| Sample metadata        |
+---------------------------+--------------------------+--------+------------------------+
|xcms fillChromPeaks        |dataMatrix.tsv            | Tabular| Data matrix            |
+---------------------------+--------------------------+--------+------------------------+
|CAMERA.annotate            |variableMetadata.tsv      | Tabular| Variable metadata      |
+---------------------------+--------------------------+--------+------------------------+



**Examples of downstream tools**

+---------------------------+--------------------------------------+--------+
| Name                      | Output file                          | Format |
+===========================+======================================+========+
|Hierarchical Clustering    |selected_metabolites_transpo.tsv      | Tabular|
+---------------------------+--------------------------------------+--------+
|ANOVA                      |selected_metabolites_transpo.tsv      | Tabular|
+---------------------------+--------------------------------------+--------+



**General schema of the metabolomic workflow**

.. image:: MetaboAnalysisCorrelation_workflow.png

-----------
Input files
-----------

+--------------------------------+------------+
| Parameter: label               | Format     |
+================================+============+
| Data matrix                    | Tabular    |
+--------------------------------+------------+
| Sample metadata                | Tabular    |
+--------------------------------+------------+
| Variable metadata              | Tabular    |
+--------------------------------+------------+
| User table file                | Tabular    |
+--------------------------------+------------+


----------
Parameters
----------

**Choice of your input files**

 | **variableMetadata**
 |
 | For example, the "variableMetadata.tsv" tabular file generated by the CAMERA.annotate step of the workflow.
 | This table must contain in particular two columns named "**pcgroup**" and "**rt**" (it is case-sensitive).
 |
 | **dataMatrix**
 |
 | For example, the "dataMatrix.tsv" tabular file generated by the CAMERA.annotate step of the workflow.
 |
 | **sampleMetadata**
 |
 | For example, the tabular file with the samples metadata generated by the xcmsSet step: one sample per line and at least two columns: ids and one variable.
 |
 | **user table**
 |
 | Tabular containing intensities where your variables (metabolites) are in columns (for example a transposition of your datamatrix file)

**Correlation threshold for pcgroup** *(metabolomic workflow only)*

The threshold value that will determine if two metabolites are correlated inside a same pcgroup after the creation of the global correlation matrix.
If you do not want to use the intra-pcgroup filter (see "corrdel" function in the description section), put this threshold to 1 and all ions will be kept.

**Choice of the correlation method**

Choose the correlation method (pearson, kendall or spearman).

**Cytoscape correlation threshold**

Choose a threshold value for selecting edges (i.e. correlations between metabolites) that will be exported to the Cytoscape sif format file.

------------
Output files
------------



**sorted_variableMetadata.tsv** *(metabolomic workflow only)*

 | A tabular file which:
 | 1) contains the original variable metadata columns
 | 2) is sorted by the pcgroup column
 | 3) contains a new column "signal_moy" (mean of all the signal values of the metabolites by sample)
 | 4) (depending of parameters) contains a "suppress" column

**correlation_matrix_selected.tsv** *(metabolomic workflow only)*

 | A correlation matrix containing only the metabolites selected in each pcgroup (metabolites tagged as "DEL" in "suppress" column are removed),
 | completed with two columns "rtmed" and "signal_moy".

**sif_table.tsv**

 | A tabular file (three columns: Metabolite1, Correlation coefficient, Metabolite 2) that can be used in Cytoscape.

------

.. class:: infomark

The output **selected_metabolites_dataMatrix.tsv** is a tabular file. You can continue your analysis using it for example in the following statistical tools:
 | Hierarchical Clustering
 | ANOVA


---------------------------------------------------

Changelog/News
--------------


**Version 1.0.1+galaxy0 - 10/12/2020**

- Update of some of the outputs' formats to match standard W4M table format
- Standard output (stdout) log improvement
- Change of testing data for faster job running for tests
- Fix: generation of the outputs for "Your table file" option

**Version 1.0.1 - 20/09/2016**

- TEST: refactoring to pass functional test using conda dependencies
- Help improvement


**Version 20141118 - 18/11/2014**

    </help>
    <citations>
        <citation type="doi">10.1093/bioinformatics/btu813</citation>
    </citations>

</tool>