Skip to content

fgiacomoni/metabolis

Repository files navigation

metabolis: An R package for metabolomics data analysis

Description

This package provides methods to perform the statistical analysis of metabolomics datasets. These methods include the reading of datasets (as 3 table dataMatrix, sampleMetadata and variableMetadata .tsv files) into an ExpressionSet object (metRead), quality control (metView) and transformation (metTransform) of the dataMatrix, and univariate hypothesis testing (metTest). Multivariate analysis and feature selection can be further performed with the ropls and biosigner packages, respectively (see the sacurine vignette).

Author

Etienne A. Thevenot (https://etiennethevenot.pagesperso-orange.fr/)

Methods

Formats

3 tabular file format used for import/export

Input (i.e. preprocessed) data consists of a 'samples times variables' matrix of intensities (datMatrix numeric matrix), in addition to sample and variable metadata (sampleMetadata and variableMetadata data frames). Theses 3 tables can be conveniently imported to/exported from R as tabular files:

ExpressionSet class used within the data analysis workflow

Within the R workflow, the ExpressionSet class perfectly handles these 3 tables (for additional information about ExpressionSet class, see the 'An introduction to Biobase and ExpressionSets' documentation from the Biobase package).

Installation

install.packages("devtools", dep=TRUE)  
devtools::install_github("https://github.com/ethevenot/metabolis")

Tutorial

See the metabolis vignette for a detailed example of the analysis of a metabolomics dataset.

Acknowledgements

This package was developed within the Metabolomics Data Sciences and Integration team at CEA, including Natacha Lenuzza, Pierrick Roger, Philippe Rinaudo, Alexis Delabriere, Camille Roquencourt, Alyssa Imbert.

Interactions with experiments from the Drug Metabolism Research Laboratory were critical to develop optimal methods for quality control and normalization, including Aurelie Roux, Samia Boudah, Florence Castelli, Benoit Colsch, Christophe Junot, Francois Fenaille.

Discussions with bioinformaticians and biostatisticians from the MetaboHUB infrastructure for metabolomics and fluxomics, and the Workflow4metabolomics Galaxy project were also of high value, including Marie Tremblay-Franco, Jean-Francois Martin, Melanie Petera, Yann Guitton, Gildas Le Corguille, Christophe Caron, Franck Giacomoni, Fabien Jourdan, Dominique Rolin.

About

Metabolomics Data Analysis

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages