MetaboDynamics : developmental version

MetaboDynamics provides a framework of Bayesian models for robust and easy analysis of longitudinal metabolomics Data. It takes concentration tables and KEGG IDs of metabolites as input and provides robust estimation of mean concentrations, functional enrichment analysis employing the KEGG database and comparison of clusters of metabolite dynamics patterns (“dynamics clusters”) under different biological conditions.

Installation

MetaboDynamics was build on R 4.4 cran. You can install the development version of MetaboDynamics from GitHub with:

if (!require("devtools", quietly = TRUE))
    install.packages("devtools")

devtools::install_github("KatjaDanielzik/MetaboDynamics")

or from Bioconductor with:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

# The following initializes usage of Bioc devel
BiocManager::install(version='devel')

BiocManager::install("MetaboDynamics")

Overview

MetaboDynamics facilitates the analysis of longitudinal metabolomics data. Common tools mostly only allow the comparison between two time points or experimental conditions and are using frequentist statistical methods.

As metabolomics data is often noisy, robust methods for the estimation of dynamics are needed. MetaboDynamics allows longitudinal analysis over multiple time points and experimental conditions employing 3 probabilistic models:

1. A hierarchical Bayesian model for the robust estimation of means at every time point despite varying spread between time points. Its outputs are A) differences between time points for every metabolite, and B) metabolite specific dynamics profiles that can be used for clustering.

2. Over-representation analysis of KEGG-functional modules in dynamics clusters with a quantitative model that employs a hypergeometric distribution and reports probabilities of a functional module being over-represented in a cluster. Can also estimate under-representation of functional modules.

3. Estimation of the distances between dynamics clusters under different experimental conditions with a Bayesian model that infers the mean pairwise Euclidean distance between two clusters. In combination with the comparison of metabolites that compose two clusters this allows to spot differences and similarities between experimental conditions.

Workflow

For a worked example see Vignette (“devel” branch: folder /vignettes, file MetaboDynamics.html) or if package is installed:

browseVignettes("MetaboDynamics")