MCMC.OTU: Bayesian analysis of multivariate counts data

This package implements poisson-lognormal generalized linear mixed model analysis of multivariate counts data using MCMC, aiming to infer the changes in relative proportions of individual variables. The package was originally designed for sequence-based analysis of microbial communities ("metabarcoding", variables = operational taxonomic units, OTUs), but can be used for other types of multivariate counts, such as in ecological applications (variables = species). The results are summarized and plotted using ggplot2 functions. Includes functions to remove sample and variable outliers and reformat counts into normalized log-transformed values for correlation and principal component/coordinate analysis. Walkthrough and examples:

Version: 1.0.8
Depends: MCMCglmm, ggplot2
Published: 2014-05-05
Author: Mikhail V. Matz
Maintainer: Mikhail V. Matz <matz at>
License: GPL-3
NeedsCompilation: no
CRAN checks: MCMC.OTU results


Reference manual: MCMC.OTU.pdf
Package source: MCMC.OTU_1.0.8.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X Snow Leopard binaries: r-release: MCMC.OTU_1.0.8.tgz, r-oldrel: MCMC.OTU_1.0.8.tgz
OS X Mavericks binaries: r-release: MCMC.OTU_1.0.8.tgz