OTUsummary {MCMC.OTU}

Summarizes and plots results of mcmc.otu() function series.

Description

Calculates abundances of each OTU across factor combinations; calculates pairwise differences between all factor combinations and their significances for each OTU; plots results as bar or line graphs with credible intervals (ggplot2) NOTE: only works for experiments involving a single multi-level fixed factor or two fully crossed multi-level fixed factors.

Usage

OTUsummary(model, data, otus = NA, relative = FALSE, 
log.base = 10, summ.plot = TRUE, ptype = "z", xgroup=NULL, ...)

Arguments

model

Model generated by mcmc.otu() or mcmc.otu.normalized()

data

Dataset used to build the model (returned by otuStack() or otuStackNormalize())

otus

A vector of OTU names to summarize and plot. If left unspecified, all OTUs will be summarized.

relative

Whether to plot OTU abundances as log(proportion of total) (default) or fold- changes relative to the sample that is considered to be "global control" (relative = TRUE). The "global control" is the combination of factors that served as a reference during model fitting, either because it is alphanumerically first (that happens by default) or because it has been explicitly designated as such using relevel() function.

log.base

Base of the logarithm to use.

summ.plot

By default, the function generates a summary plot, which is a line-points-95% credible intervals plot of log(fraction of total) with 'relative=FALSE' and a bar graph of log(fold change relative to the control), again with 95% credible intervals, with 'relative=TRUE'. Specify 'summ.plot=FALSE' if you don't want the summary plot.

ptype

Which type of p-values to use. By default p-values based on the Bayesian z-score are used. Specify 'ptype="mcmc"' to output more conventional p-values based on MCMC sampling (these will be limited on the lower end by the size of MCMC sample).

xgroup

For two-factor designs: which of the factors to use to form the x-axis. The other one will be used to form facets.

...

Additional options for summaryPlotOTU() function. Among those, 'x.order' can be a vector specifying the order of factor levels on the x-axis.

Value

A list of three items:

summary

Summary table containing calculated abundances, their SD and 95% credible limits

otuWise

A series of matrices listing pairwise differences between factor combinations (upper triangle) and corresponding p-values (lower triangle)

ggPlot

the ggplot2 object for plotting. See http://docs.ggplot2.org/0.9.2.1/theme.html for ways to modify it, such as add text, rotate labels, change fonts, etc.

Author(s)

Mikhail V. Matz, University of Texas at Austin <[e-mail]>

References

Elizabeth A. Green, Sarah W. Davies, Mikhail V. Matz, Monica Medina Next-generation sequencing reveals cryptic Symbiodinium diversity within Orbicella faveolata and Orbicella franksi at the Flower Garden Banks, Gulf of Mexico. PeerJ 2014 https://peerj.com/preprints/246/

See Also

mcmc.otu(),MCMCglmm()

Examples

# see example in ?MCMC.OTU