Mercurial > repos > eschen42 > w4mkmeans
diff w4mkmeans.xml @ 2:c415b7dc6f37 draft default tip
planemo upload for repository https://github.com/HegemanLab/w4mkmeans_galaxy_wrapper/tree/master commit 3e916537da6bb37e6f3927d7a11e98e0ab6ef5ec
| author | eschen42 |
|---|---|
| date | Mon, 05 Mar 2018 12:40:17 -0500 |
| parents | 02cafb660b72 |
| children |
line wrap: on
line diff
--- a/w4mkmeans.xml Wed Aug 09 18:06:55 2017 -0400 +++ b/w4mkmeans.xml Mon Mar 05 12:40:17 2018 -0500 @@ -1,9 +1,11 @@ -<tool id="w4mkmeans" name="w4mKmeans" version="0.98.3"> - <description>Calculate K-means for W4M dataMatrix features or samples</description> +<tool id="w4mkmeans" name="Kmeans for W4m" version="0.98.4"> + <description>Calculate K-means for W4m dataMatrix features or samples</description> <requirements> - <requirement type="package" version="3.3.2">r-base</requirement> + <requirement type="package" version="3.4.1">r-base</requirement> <requirement type="package" version="1.1_4">r-batch</requirement> + <requirement type="package" version="1.8.0">libssh2</requirement> + <requirement type="package" version="1.13.2">krb5</requirement> </requirements> <stdio> @@ -27,19 +29,18 @@ slots "\${GALAXY_SLOTS:-1}" variableMetadata_out '$variableMetadata_out' variable_metadata_path '$variableMetadata_in' - ; echo exit code $? ]]></command> <inputs> - <param name="dataMatrix_in" label="Data matrix file" type="data" format="tabular" help="variable x sample, decimal: '.', missing: NA, mode: numerical, separator: tab" /> - <param name="sampleMetadata_in" label="Sample metadata file" type="data" format="tabular" help="sample x metadata columns, separator: tab" /> - <param name="variableMetadata_in" label="Variable metadata file" type="data" format="tabular" help="variable x metadata columns, separator: tab" /> - <param name="categoricalPrefix" label="prefix for cluster names " type="text" value="k" help="[categorical_prefix] Some tools require non-numeric values to discern categorical data; e.g., enter 'k' here to prepend 'k' to cluster numbers in the output; default 'k'." /> + <param name="dataMatrix_in" label="Data matrix file" type="data" format="tabular" help="Feature (variable) x sample; decimal point: '.'; missing: NA; mode: numerical; separator: tab" /> + <param name="sampleMetadata_in" label="Sample metadata file" type="data" format="tabular" help="Sample x metadata columns; separator: tab" /> + <param name="variableMetadata_in" label="Variable (feature) metadata file" type="data" format="tabular" help="Feature (variable) x metadata columns; separator: tab" /> + <param name="categoricalPrefix" label="Prefix for cluster names " type="text" value="c" help="String prepended to cluster numbers in output; default 'c'; leave blank for no prefix." /> <param name="ksamples" label="K value(s) for samples" type="text" value = "0" help="[ksamples] Single K or comma-separated Ks for samples, or 0 for none." /> <param name="kfeatures" label="K value(s) for features" type="text" value = "0" help="[kfeatures] Single K or comma-separated Ks for features (variables), or 0 for none." /> - <param name="iter_max" label="Max number of iterations" type="text" value = "10" help="[iter_max] The maximum number of iterations allowed; default 10." /> - <param name="nstart" label="Number of random sets" type="text" value = "1" help="[nstart] How many random sets should be chosen; default 1." /> - <param name="algorithm" label="Algorithm for clustering" type="select" value = "Hartigan-Wong" help="[algorithm] K-means clustering algorithm, default 'Hartigan-Wong'; alternatives 'Lloyd', 'MacQueen'; 'Forgy' is a synonym for 'Lloyd', see stats::kmeans reference for further info."> + <param name="iter_max" label="Maximum number of iterations" type="text" value = "20" help="[iter_max] The maximum number of iterations allowed; default 20." /> + <param name="nstart" label="Number of random sets of clusters" type="text" value = "20" help="[nstart] How many random sets of clusters should be chosen initially; default 20." /> + <param name="algorithm" label="Algorithm for clustering" type="select" value = "Hartigan-Wong" help="[algorithm] K-means clustering algorithm, default 'Hartigan-Wong'; <br />alternatives 'Lloyd', 'MacQueen'; 'Forgy' (synonym for 'Lloyd'); see references."> <option value="Forgy">Forgy</option> <option value="Hartigan-Wong" selected="True">Hartigan-Wong</option> <option value="Lloyd">Lloyd</option> @@ -48,9 +49,9 @@ </inputs> <outputs> - <data name="sampleMetadata_out" label="${tool.name}_${sampleMetadata_in.name}" format="tabular" ></data> - <data name="variableMetadata_out" label="${tool.name}_${variableMetadata_in.name}" format="tabular" ></data> - <data name="scores_out" label="${tool.name}_${dataMatrix_in.name}.kmeans" format="tabular" ></data> + <data name="sampleMetadata_out" label="${sampleMetadata_in.name}.kmeans-smpl" format="tabular" ></data> + <data name="variableMetadata_out" label="${variableMetadata_in.name}.kmeans-vrbl" format="tabular" ></data> + <data name="scores_out" label="${dataMatrix_in.name}.kmeans-score" format="tabular" ></data> </outputs> <tests> @@ -60,17 +61,17 @@ <param name="variableMetadata_in" value="input_variableMetadata.tsv"/> <param name="ksamples" value="3,4"/> <param name="kfeatures" value="5,6,7"/> - <param name="iter_max" value="10"/> - <param name="nstart" value="1"/> + <param name="iter_max" value="20"/> + <param name="nstart" value="20"/> <param name="algorithm" value="Hartigan-Wong"/> <output name="scores_out"> <assert_contents> <has_text text="proportion" /> <has_text text="0.87482" /> - <has_text text="0.89248" /> - <has_text text="0.95355" /> - <has_text text="0.95673" /> - <has_text text="0.95963" /> + <has_text text="0.91765" /> + <has_text text="0.95362" /> + <has_text text="0.95719" /> + <has_text text="0.97966" /> </assert_contents> </output> </test> @@ -79,57 +80,38 @@ <help> <![CDATA[ +=========================== +K-means for W4m data matrix +=========================== + **Author** - Arthur Eschenlauer (University of Minnesota, esch0041@umn.edu) ---------------------------------------------------------------------------- - - -**Source** - The source code for the w4mkmeans tool is available (from the Hegeman lab github repository) at https://github.com/HegemanLab/w4mkmeans_galaxy_wrapper - -**R code used** - The R code invoked by this wrapper is the R 'stats::kmeans' package - ----------------------------------------------------------------------------------------------------------------------------------------------------------------------- - +**Source Code** - The source code for the w4mkmeans tool is available (from the Hegeman lab github repository) at https://github.com/HegemanLab/w4mkmeans_galaxy_wrapper -**Tool updates** - -See the **NEWS** section at the bottom of this page +**R code used** - The R code invoked by this wrapper is the R kmeans package at https://stat.ethz.ch/R-manual/R-devel/library/stats/html/kmeans.html ---------------------------------------------------- - -=========================== -K-means for W4M data matrix -=========================== +**Tool updates** - See the **NEWS** section at the bottom of this page ----------- Description ----------- -Calculate K-means for sample-clusters (or feature-clusters, or both) using W4M dataMatrix (i.e., XCMS-preprocessed data files) as input. +This tool calculate K-means clusters for samples, features, or both using W4m dataMatrix (i.e., XCMS-preprocessed data files) as input and writes the results to new columns in sampleMetadata, variableMetadata, or both. -*Please note that XCMS refers to features as 'variables'. This documentation does not use either term consistently.* +- If several, comma-separated K's are supplied, then one column is added for each K. +- For feature-clustering, each feature is assigned to a cluster such that the feature's response for all samples is closer to the mean of all features for that cluster than to the mean for any other cluster. +- For sample-clustering, each sample is assigned to a cluster such that the sample's response for all features is closer to the mean of all samples for that cluster than to the mean for any other cluster. +- Clustering is mutually exclusive, **not** hierarchical. + - Hierarchical clustering is available through the W4m Heat Map tool, https://github.com/workflow4metabolomics/heatmap ----------------- Workflow Position ----------------- - - Tool category: Statistical Analysis - - Upstream tool category: Preprocessing - - Downstream tool categories: Statistical Analysis - - ----------- -Motivation ----------- - -This tool clusters samples, features (variables), or both from the W4M dataMatrix and writes the results to new columns in sampleMetadata, variableMetadata, or both, respectively. - - - If several, comma-separated K's are supplied, then one column is added for each K. - - This clustering is **not** hierarchical; each member of a cluster is not a member of any other cluster. - - For feature-clustering, each feature is assigned to a cluster such that the feature's response for all samples is closer to the mean of all features for that cluster than to the mean for any other cluster. - - For sample-clustering, each sample is assigned to a cluster such that the sample's response for all features is closer to the mean of all samples for that cluster than to the mean for any other cluster. - +- Tool category: Statistical Analysis +- Upstream tool category: Preprocessing +- Downstream tool categories: Statistical Analysis ----------- Input files @@ -152,92 +134,117 @@ **Data matrix** - input-file dataset - - XCMS variable x sample 'dataMatrix' (tabular separated values) file of the numeric data matrix, with . as decimal, and NA for missing values; the table must not contain metadata apart from row and column names; the row and column names must be identical to the rownames of the sample and feature metadata, respectively (see below) +- W4m variable (i.e. feature) x sample 'dataMatrix' (tabular separated values) file of the numeric data matrix, with . as decimal, and NA for missing values; the table must not contain metadata apart from row and column names; the row and column names must be identical to the rownames of the sample and feature metadata, respectively (see below) **Sample metadata** - input-file dataset - - XCMS sample x metadata 'sampleMetadata' (tabular separated values) file of the numeric and/or character sample metadata, with . as decimal and NA for missing values +- W4m sample x metadata 'sampleMetadata' (tabular separated values) file of the numeric and/or character sample metadata, with . as decimal and NA for missing values **Feature metadata** - input-file dataset - - XCMS variable x metadata 'variableMetadata' (tabular separated values) file of the numeric and/or character feature metadata, with . as decimal and NA for missing values +- W4m variable (i.e. feature) x metadata 'variableMetadata' (tabular separated values) file of the numeric and/or character feature metadata, with . as decimal and NA for missing values -**kfeatures** - K or K's for features (default = 0) +**Prefix for cluster names** - character(s) to add as prefix to category number (default = 'c') + +- Some tools treat only non-numeric data as categorical; this prefix ensures that clusters data will be treated as categorical; an empty string is permitted here if desired (and succeeding tools requiring categorical data accept integers). - - integer or comma-separated integers ; zero (the default) or less will result in no calculation. +**K-values for samples** - K or K-range for samples (default = 0) -**ksamples** - K or K-range for samples (default = 0) +- Integer or comma-separated positive integers ; zero (or less) will result in no calculation. - - integer or comma-separated integers ; zero (the default) or less will result in no calculation. +**K-values for features** - K or K's for features (default = 0) -**iter_max** - maximum_iterations (default = 10) +- Integer or comma-separated positive integers ; zero (or less) will result in no calculation. - - maximum number of iterations per calculation (see https://stat.ethz.ch/R-manual/R-devel/library/stats/html/kmeans.html). +**Maximumn number of iterations** - (default = 20) -**nstart** - how many random sets should be chosen (default = 1) +- Maximum number of iterations per calculation (see https://stat.ethz.ch/R-manual/R-devel/library/stats/html/kmeans.html for further info). - - maximum number of iterations per calculation (see https://stat.ethz.ch/R-manual/R-devel/library/stats/html/kmeans.html). +**Number of random sets** - how many random sets should be chosen to start (default = 20) + +- Number of random sets of clusters to be chosen to start calculation (see https://stat.ethz.ch/R-manual/R-devel/library/stats/html/kmeans.html for further info). -**categorical_prefix** - character(s) to add as prefix to category number (default = 'k') +**Algorithm** - Algorithm for clustering" (default = Hartigan-Wong) - - some tools treat only non-numeric data as categorical; this prefix ('k' by default) ensures that clusters data will be treated as categorical; an empty string is permitted here if desired (and succeeding tools accept integers as categorical data). +- K-means clustering algorithm: 'Hartigan-Wong', 'Lloyd', or 'MacQueen'; 'Forgy' is a synonym for 'Lloyd' (see references for further info). ------------ Output files ------------ -**XCMS sampleMetadata** - (tabular separated values) file identical to the Sample metadata file given as an input argument, excepting one column added for each K +**Sample Metadata** - (tabular separated values) file identical to the Sample metadata file given as an input argument, excepting one column added for each K - - **k#** - cluster number for clustering samples with K = # +- **k#** - cluster number for clustering samples with K = # -**XCMS variableMetadata** - (tabular separated values) file identical to the Feature metadata file given as an input argument, excepting one column added for each K +**Variable Metadata** - (tabular separated values) file identical to the Feature metadata file given as an input argument, excepting one column added for each K - - **k#** - cluster number for clustering features with K = # +- **k#** - cluster number for clustering features with K = # **scores** - (tabular separated values) file with one line for each K. - - **clusterOn** - what was clustered - either 'sample' or 'feature' - - **k** - the chosen K for clustering - - **totalSS** - total (*between-treatements* plus total of *within-treatements*) sum of squares - - **betweenSS** - *between-treatements* sum of squares - - **proportion** - betweenSS / totalSS +- **clusterOn** - what was clustered - either 'sample' or 'feature' +- **k** - the chosen K for clustering +- **totalSS** - total (*between-treatements* plus total of *within-treatements*) sum of squares +- **betweenSS** - *between-treatements* sum of squares +- **proportion** - betweenSS / totalSS --------------- Working example --------------- +.. class:: infomark + **Input files** -+-------------------+-------------------------------------------------------------------------------------------------------------------+ -| Input File | Download from URL | -+===================+===================================================================================================================+ -| Data matrix | https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/input_dataMatrix.tsv | -+-------------------+-------------------------------------------------------------------------------------------------------------------+ -| Sample metadata | https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/input_sampleMetadata.tsv | -+-------------------+-------------------------------------------------------------------------------------------------------------------+ -| Feature metadata | https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/input_variableMetadata.tsv | -+-------------------+-------------------------------------------------------------------------------------------------------------------+ ++-------------------------------------------------------------------------------------------------------------------+ +| URL | ++===================================================================================================================+ +| https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/input_dataMatrix.tsv | ++-------------------------------------------------------------------------------------------------------------------+ +| https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/input_sampleMetadata.tsv | ++-------------------------------------------------------------------------------------------------------------------+ +| https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/input_variableMetadata.tsv | ++-------------------------------------------------------------------------------------------------------------------+ + +.. class:: infomark **Other input parameters** +-----------------+---------------+ | Input Parameter | Value | +=================+===============+ +| prefix | c | ++-----------------+---------------+ | ksamples | 3,4 | +-----------------+---------------+ | kfeatures | 5,6,7 | +-----------------+---------------+ -| iter_max | 10 | +| iter_max | 20 | +-----------------+---------------+ -| nstart | 1 | +| nstart | 20 | +-----------------+---------------+ | algorithm | Hartigan-Wong | +-----------------+---------------+ +.. class:: infomark + +**Expected output files** + ++-------------------------------------------------------------------------------------------------------------------+ +| URL | ++===================================================================================================================+ +| https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/output_kmeans-score.tsv | ++-------------------------------------------------------------------------------------------------------------------+ +| https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/output_kmeans-vrbl.tsv | ++-------------------------------------------------------------------------------------------------------------------+ +| https://raw.githubusercontent.com/HegemanLab/w4mkmeans_galaxy_wrapper/master/test-data/output_kmeans-smpl.tsv | ++-------------------------------------------------------------------------------------------------------------------+ + ---- NEWS ---- +- February 2018, Version 0.98.4 - Renamed output datasets to append '``.kmeans-smpl``', '``.kmeans-vrbl``', or '``.kmeans-score``'; refactored multi-threading. - August 2017, Version 0.98.3 - Add (optional) prefix to category numbers for downstream tools that treat only non-numeric data as categorical. - August 2017, Version 0.98.1 - First release @@ -277,28 +284,12 @@ year = 1965 } ]]></citation> - <!-- W4M 3.0 - Guitton et al. 2017--> + <!-- W4m 3.0 - Guitton et al. 2017--> <citation type="doi">10.1016/j.biocel.2017.07.002</citation> - <!-- W4M 2.5 - Giacomini et al. 2014 --> + <!-- W4m 2.5 - Giacomini et al. 2014 --> <citation type="doi">10.1093/bioinformatics/btu813</citation> <!-- Hartigan and Wong algorithm --> - <citation type="bibtex"><![CDATA[ -@article{Hartigan79, - added-at = {2007-02-27T16:22:09.000+0100}, - author = {Hartigan, J. and Wong, M.}, - biburl = {https://www.bibsonomy.org/bibtex/23d8bfc440c5725783876929c022f67ce/pierpaolo.pk81}, - description = {WSD}, - interhash = {10d6d33920d9af578a4d0a556dc1477d}, - intrahash = {3d8bfc440c5725783876929c022f67ce}, - journal = {Applied Statistics}, - keywords = {imported}, - pages = {100-108}, - timestamp = {2007-02-27T16:22:11.000+0100}, - title = {Algorithm AS136: A k-means clustering algorithm}, - volume = 28, - year = 1979 -} - ]]></citation> + <citation type="doi">10.2307/2346830</citation> <!-- Lloyd algorithm --> <citation type="doi">10.1109/TIT.1982.1056489</citation> <!-- MacQueen algorithm -->