view rgcca.xml @ 1:4e73ea176c34 draft default tip

"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/rgcca commit ce05b5eb018ae1c4d580ab5ce1a33896c1aa8c5b"

<tool id="rgcca" name="RGCCA" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@">

    <description>performs multiblock data analysis of several sets of variables (blocks) observed on the same group of individuals.</description>

    <macros>
        <import>macro.xml</import>
    </macros>

    <edam_topics>
        <edam_topic>topic_2269</edam_topic>
    </edam_topics>

    <edam_operations>
        <edam_operation>operation_2945</edam_operation>
        <edam_operation>operation_3465</edam_operation>
        <edam_operation>operation_0337</edam_operation>
    </edam_operations>

    <requirements>
        <requirement type="package" version="@TOOL_VERSION@">rgccacmd</requirement>
        <requirement type="package" version="4.1">r-base</requirement>
    </requirements>

    <command detect_errors="exit_code"><![CDATA[
        #set data_paths = ",".join([str(_.file_name) for _ in $blocks])
        #set data_names = ",".join([str(_.element_identifier).replace(',', '_') for _ in $blocks])
        #set out_files = str($output_selector).split(",")
        Rscript '$__tool_directory__/launcher.R'
            --datasets '${data_paths}'
            --names '${data_names}'
            #if 'individuals' in $out_files
                --o1 '$individual_plot'
            #end if
            #if 'corcircle' in $out_files
                --o2 '$corcircle'
            #end if
            #if 'top_variables' in $out_files
                --o3 '$top_variables'
            #end if
            #if 'ave' in $out_files
                --o4 '$ave'
            #end if
            #if 'design' in $out_files
                --o5 '$design'
            #end if
            #if 'individual_table' in $out_files
                --o6 '$individual_table'
            #end if
            #if 'variable_table' in $out_files
                --o7 '$variable_table'
            #end if
            #if 'rdata' in $out_files
                --o8 '$rdata'
            #end if
            $parse.header
            --separator $parse.separator
            $analyse.superblock
            $analyse.scale
            #if $analyse.tau.bool == 'false'
                --penalty $analyse.tau.value
            #else
                --penalty $analyse.tau.bool
            #end if
            --ncomp $analyse.ncomp
            --scheme $analyse.scheme
            #if $analyse.method.family == '1'
                --type pca
            #else
                --type $analyse.method.type
            #end if
            #if $analyse.connection
                --connection $analyse.connection
            #end if
            #if $analyse.supervised.learning_mode == 'supervised'
                --response $analyse.supervised.block_response
            #end if
            #if $graphic.response
                --group $graphic.response
            #end if
            --compx $graphic.compx
            --compy $graphic.compy
            --nmark $graphic.nmark
            $graphic.text
            --block $graphic.blockx
            --block_y $graphic.blocky
    ]]></command>

    <inputs>
        <param name="blocks" type="data" format="tsv,tabular,txt,csv" multiple="true" optional="false" label = "Load blocks"
            help="TSV file containing a matrix with: (i) quantitative values only (decimal should be separated by '.'), (ii) the samples in lines (should be labelled in the 1rst column) and (iii) variables in columns (should have a header)."/>

        <section name="parse" title="Advanced parsing" help="By default, on tabulated files with a header.">
            <param name="header" type="boolean" truevalue="" falsevalue="-H" checked="true" label="Consider the first row as header of columns" help="Used for both blocks and color files."/>
            <param name="separator" type="select" display="radio" label="Column separator" help="Character used to separate the column (for all blocks, connection and color files).">
                <option value="1" selected="true">Tabulation</option>
                <option value="2">Semicolon</option>
            </param>
        </section>

        <section name="analyse" title="Advanced analysis"
            help="By default, the analysis: is a Regularised Generalised Canonical Correlation Analysis, scales the blocks, uses a superblock with a factorial scheme function, a tau equals to one and two components for each block.">

            <param name="ncomp" type="integer" label="Number of component" value="2" min="2" max="5"
                help="The number of component to use in the analysis for each block (should not be greater than the minimum number of variable among the blocks)."/>

            <param name="scale" type="boolean" truevalue="" falsevalue="--scale" checked="true" label="Scale the blocks"
                help="A data centering step is always performed. If activated, each block is normalised and divided by the square root of its number of variables."/>

            <conditional name="method">

                <param name="family" type="select" label="Analysis method">
                    <option value="1">One block</option>
                    <option value="2">Two blocks</option>
                    <option value="m" selected="true">Multiple blocks</option>
                    <option value="ms">Multiple blocks with superblock</option>
                </param>

                <when value="2">
                    <param name="type" type="select" label=" ">
                        <option value="pls">Partial Least Squares Regression</option>
                        <option value="cca">Canonical Correlation Analysis</option>
                        <option value="ifa">Interbattery Factor Analysis</option>
                        <option value="ra">Redundancy analysis</option>
                    </param>
                </when>

                <when value="m">
                    <param name="type" type="select" label=" ">
                        <option value="rgcca">Regularized Generalized CCA</option>
                        <option value="sgcca">Sparse Generalized CCA</option>
                        <option value="sumcor">SUM of CORrelations method</option>
                        <option value="ssqcor">Sum of SQuared CORrelations method</option>
                        <option value="sabscor">Sum of ABSolute value CORrelations method</option>
                        <option value="sumcov">SUM of COVariances method</option>
                        <option value="ssqcov">Sum of SQuared COVariances method</option>
                        <option value="sabscov">Sum of ABSolute value COVariances method</option>
                        <option value="maxbet">MAXBET</option>
                        <option value="maxbet-b">MAXBET-B</option>
                    </param>
                </when>

                <when value="ms">
                    <param name="type" type="select" label=" ">
                        <option value="gcca">Generalized CCA</option>
                        <option value="hpca">Hierarchical PCA</option>
                        <option value="mfa">Multiple Factor Analysis</option>
                    </param>
                </when>

                <when value="1"/>

            </conditional>

            <param name="connection" optional="true" type="data" format="tsv,tabular,txt,csv" label="Load the design matrix (if superblock or supervised disabled)"
                help="TSV file without header and without row names. This file describes the connections between the blocks. It should contain 1 (if two blocks are related) or 0 values otherwise. The columns are separated by tabulations. It is a symmetric matrix with the same dimension as the number of blocks."/>

            <param name="superblock" type="boolean" truevalue="" falsevalue="--superblock" checked="true" label="Use a superblock"
                help="A block defined as the concatenation of all the other blocks. The space spanned by global components is viewed as a compromise space that integrated all the modalities and facilitates the visualization of the results and their interpretation. If disabled, all blocks are assumed to be connected or a connection file could be used."/>

            <conditional name="supervised">
                <param name="learning_mode" type="select" display="radio" label="Learning mode">
                    <option value="unsupervised">Unsupervised</option>
                    <option value="supervised">Supervised</option>
                </param>
                <when value="supervised">
                    <param name="block_response" type="integer" value="1" min="1" max="10" label="Use a block as response (supervised analysis)" help="@BLOCK_RULES@ By default, the first block is selected."/>
                </when>
                <when value="unsupervised"/>
            </conditional>

            <conditional name="tau">
                <param name="bool" type="select" display="radio" label="Tau selection"
                    help="For RGCCA, a regularization parameter for each block (i.e., tau) [default: 1]. Tau varies from 0 (maximizing the correlation) to 1 (maximizing the covariance). For SGCCA, tau is automatically set to 1 and a shrinkage parameter can be defined instead for automatic variable selection, varying from the square root of the variable number (the fewest selected variables) to 1 (all the variables are included).">
                    <option value="false">Manual</option>
                    <option value="optimal">Optimal</option>
                </param>
                <when value="false">
                    <param name="value" type="float" label=" " value="1" min="0" max="1"/>
                </when>
                <when value="optimal"/>
            </conditional>

            <param name="scheme" type="select" label="Scheme function" help="Link (i.e. scheme) function for covariance maximization is calculated with: the identity function (horst scheme),
the absolute values (centroid scheme), the squared values (factorial scheme). Only, the horst scheme penalizes structural
negative correlation. The factorial scheme discriminates more strongly the blocks than the centroid one.">
                <option value="1">Horst : f(x)</option>
                <option value="2" selected="true">Factorial : f(x)^2</option>
                <option value="3">Centroid : f|x|</option>
                <option value="4">Other: f(x)^4</option>
            </param>

        </section>

        <section name="graphic" title="Advanced graphic" help="By default, the x-axis and y-axis are respectively the first and the second components, the number of top variables is 100 and a superblock is used.">
            <param name="response" optional="true" type="data" format="tsv,tabular,txt,csv" label="Color the individual plot with a response variable"
                help="A TSV file containing either: (i) an only column with a qualitative or a quantitative variable; (ii) multiple columns corresponding to a disjunctive table."/>
            <param name="text" type="boolean" truevalue="" falsevalue="--text" checked="true" label="Display the names of the points (in biplots)"/>
            <param name="compx" type="integer" label="Component for the X-axis" help="The component used in the X-axis for biplots and the only component used for top variable plot. @COMP_RULES@" value="1" min="1" max="5"/>
            <param name="compy" type="integer" label="Component for the Y-axis" help="The component used in the Y-axis for biplots. @COMP_RULES@" value="2" min="1" max="5"/>
            <param name="blockx" type="integer" value="0" min="0" max="10" label="Visualise this block" help="Block used in the X-axis for individual plot and the only block used for corcircle and top variable plots. @BLOCK_RULES@"/>
            <param name="blocky" type="integer" value="0" min="0" max="10" label="Visualise this block for the Y-axis (in individual plot)" help="0 corresponds to the superblock (or the last block loaded), @BLOCK_RULES@ By default, the superblock is selected."/>
            <param name="nmark" type="integer" label="Number of top variables" value="100" min="10" max="300"/>
        </section>

        <param name="output_selector" type="select" multiple="true" label="Outputs">
            <option value="individuals" selected="true">Individual plot</option>
            <option value="corcircle" selected = "true">Corcircle plot</option>
            <option value="top_variables">Top variables plot</option>
            <option value="ave">Explained variance plot</option>
            <option value="design">Design plot</option>
            <option value="individual_table" selected="true">Individual table</option>
            <option value="variable_table" selected="true">Variable table</option>
            <option value="rdata">RData file</option>
        </param>
    </inputs>

    <outputs>
        <data name="individual_plot" label="${tool.name} on ${on_string}: individuals.pdf" format="pdf">
            <filter>"individuals" in output_selector</filter>
        </data>
        <data name="corcircle" label="${tool.name} on ${on_string}: corcircle.pdf" format="pdf">
            <filter>"corcircle" in output_selector</filter>
        </data>
        <data name="top_variables"  label="${tool.name} on ${on_string}: top_variables.pdf" format="pdf">
            <filter>"top_variables" in output_selector</filter>
        </data>
        <data name="ave"  label="${tool.name} on ${on_string}: ave.pdf" format="pdf">
            <filter>"ave" in output_selector</filter>
        </data>
        <data name="design"  label="${tool.name} on ${on_string}: design.pdf" format="pdf">
            <filter>"design" in output_selector</filter>
        </data>
        <data name="individual_table"  label="${tool.name} on ${on_string}: individuals.tsv" format="tsv">
            <filter>"individual_table" in output_selector</filter>
        </data>
        <data name="variable_table"  label="${tool.name} on ${on_string}: variables.tsv" format="tsv">
            <filter>"variable_table" in output_selector</filter>
        </data>
        <data name="rdata"  label="${tool.name} on ${on_string}: rgcca.result.RData" format="rdata">
            <filter>"rdata" in output_selector</filter>
        </data>
    </outputs>

    <tests>

        <test expect_num_outputs="8" expect_exit_code="0">
            <expand macro="output_tests" path="1block"/>
            <param name="blocks" value="agriculture.tsv" ftype = "tsv"/>
            <output name="individual_table">
                <assert_contents>
                    <has_n_columns n="4"/>
                    <has_line_matching
                            expression='"agriculture.component1"\s"agriculture.component2"\s"superblock.component1"\s"superblock.component2"'/>
                    <has_line_matching
                            expression='^.+(\s\-?\d+.\d+){4}$'/>
                </assert_contents>
            </output>
        </test>

        <test expect_num_outputs="8" expect_exit_code="0">
            <expand macro="output_tests" path="3blocks_connection"/>
            <expand macro="output_tests_3blocks"/>
            <section name="analyse">
                <param name="connection" value="connection.tsv" ftype = "tsv"/>
                <param name="superblock" value="false"/>
            </section>
            <assert_command>
                <has_text text="-connection"/>
                <has_text text="--superblock"/>
            </assert_command>
        </test>

        <test expect_num_outputs="8" expect_exit_code="0">
            <expand macro="output_tests" path="3blocks_supervised"/>
            <expand macro="output_tests_3blocks"/>
            <section name="analyse">
                <param name="superblock" value="false"/>
                <conditional name="supervised" >
                    <param name="learning_mode" value="supervised"/>
                    <param name="block_response" value="3"/>
                </conditional>
            </section>
            <assert_command>
                <has_text text="--response 3"/>
                <has_text text="--superblock"/>
            </assert_command>
        </test>

        <test expect_num_outputs="8" expect_exit_code="0">
            <expand macro="output_tests" path="3blocks"/>
            <expand macro="output_tests_3blocks"/>
        </test>

        <test expect_num_outputs="8" expect_exit_code="0">
            <expand macro="output_tests" path="3blocks_sgcca"/>
            <expand macro="output_tests_3blocks"/>
            <section name="analyse">
                <conditional name="method">
                    <param name="family" value="m"/>
                    <param name="type" value="sgcca"/>
                </conditional>
            </section>
            <assert_command>
                <has_text text="sgcca"/>
            </assert_command>
        </test>

        <test expect_num_outputs="8" expect_exit_code="0">
            <expand macro="output_tests" path="2blocks"/>
            <param name="blocks" value="agriculture.tsv,politic.tsv"/>
            <section name="analyse">
                <param name="scale" value="false"/>
                <conditional name="tau">
                    <param name="bool" value="false"/>
                    <param name="value" value="0"/>
                </conditional>
                <param name="scheme" value="3"/>
                <param name="ncomp" value="3"/>
                <conditional name="method">
                    <param name="family" value="2"/>
                    <param name="type" value="pls"/>
                </conditional>
            </section>
            <section name="graphic">
                <param name="response" value="political_system.tsv" ftype = "tsv"/>
                <param name="text" value="false"/>
                <param name="compx" value="3"/>
                <param name="compy" value="1"/>
                <param name="blockx" value="2"/>
                <param name="blocky" value="1"/>
                <param name="nmark" value="11"/>
            </section>
            <assert_command>
                <has_text text="pls"/>
                <has_text text="--group"/>
            </assert_command>
        </test>

    </tests>
<help>

==================================
ABOUT
==================================


**Author:**
Etienne CAMENEN


**Contact:**
[e-mail]


**R package:**
    | The RGCCA package is available from the CRAN repository (v2.1.2; https://cran.r-project.org/web/packages/RGCCA).
    | This tool is based on a version available on github (v3.0; https://github.com/rgcca-factory/RGCCA).

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

==================================
R/SGCCA
==================================

A user-friendly multi-blocks analysis (Regularized Generalized Canonical Correlation Analysis, RGCCA) as described in [1] and [2] with all default settings predefined. The software produces figures to explore the analysis' results: individuals and variables projected on two components of the multi-block analysis, list of top variables and explained variance in the model.

**Working example**

    | From Russett data (RGCCA package): https://github.com/BrainAndSpineInstitute/rgcca_ui/tree/master/inst/extdata
    | Use *agriculture.tsv* as a block. Add *industry.tsv* and *politic.tsv* for multiblock analysis. *connection.tsv* could be used as a design matrix and *political_system.tsv* as a response variable respectively in analysis and graphic settings.

**Documentation**

- RGCCA: https://cran.r-project.org/web/packages/RGCCA/vignettes/vignette_RGCCA.pdf
- accepted input / https://github.com/BrainAndSpineInstitute/rgcca_ui#input-files
- tutorial: https://github.com/BrainAndSpineInstitute/rgcca_galaxy#readme

</help>

    <citations>
        <citation type="doi">10.1007/s11336-017-9573-x</citation>
        <citation type="doi">10.1007/s11336-011-9206-8</citation>
    </citations>

</tool>