PCoA: Principal coordinates analysis

In katrikorpela/mare: Microbiota Analysis in R Easily

Principal coordinates analysis

Description

The function conducts principal coordinates analysis using Bray-Curtis dissimilarities, interpolates values of a continuous variable into the principal coordinate space, and plots the PCoA with the interpolated continuous variable as the background color. Group differences are shown by different-coloured symbols and the proportion of overall variation explained by group is printed on the bottom of the figure with a p-value from adonis in package vegan.

Usage

PCoA(taxonomic.table, meta, readcount.cutoff = 0, group = NULL, group2 = NULL,
    components = c(1, 2), background.variable = NULL, colour.scheme = "terrain.colors", 
    ellipse = F, hull = F, spider = F, legendplace = "topright", 
    select.by = NULL, select = NULL, relative = F, pdf = F, keep.result = F,
    subjectID = NULL, time = NULL, distance = "bray", constrain = NULL)

Arguments

taxonomic.table	Name of the taxonomic table. Should be the name of a text file.
meta	Name of the metadata file containing the grouping variable. Should be the name of a text file.
readcount.cutoff	Lowest acceptable read count per sample. Samples with fewer reads are ignored.
group	Name of the grouping variable.
group2	Name of a second grouping variable.
components	Which principal components should be shown? Defaults to the first two components.
background.variable	Name of the continuous variable from the metadata file or a bacterial taxon to be shown as the background colour.
colour.scheme	Backround colour scheme. Can be a vector of colour codes or names, or pre-specified colour scheme such as "terrain.colors" (default), "heat.colors","topo.colors", or "cm.colors". NOTE: These must be in quotes!
ellipse	Draw standard deviation ellipse for each group using function ordiellipse from package vegan.
hull	Draw outline for each group using function ordihull from package vegan.
spider	Draw lines joining the samples within each group using function ordispider from package vegan.
legendplace	Place of the group legend in the figure. Options are "bottomright", "bottom", "bottomleft", "left", "topleft", "top", "topright", "right" and "center".
select.by	Name of a variable in the metadata file by which a subset will be selected for plotting.
select	Determines which value of the selection variable will be selected for plotting.
relative	Should relative abundances rather than absolute counts be used? If yes, specify TRUE.
pdf	Should the figure be saved as pdf? If yes, specify TRUE.
keep.result	Should the PC-component scores be returned? If yes, specify TRUE.
subjectID	Column name for subject ID in the metadata file, if several samples from the same subjects are present and should be linked in the plot.
time	Column name for time/order variable in the metadata file, if several samples from the same subjects are present and should be linked in the plot.
distance	Use Bray-Curtis distance ("bray"), Pearson correlation distance ("pearson"), Euclidean distance ("euc").
constrain	To constrain the first principal component to be a specified variable, specify a variable name.

Value

Returns a table with correlations of the bacterial taxa with the selected principal coordinates and p-values of the correlations.

Author(s)

Katri Korpela

Examples

## Not run: 
#Plot the first two principal coordinates in the genus-level data, ignoring samples with 
#fewer than 2000 reads, showing differences between treatment groups, and the OTU 
#richness as the background colour.	Save the plot as PDF. Save the result as R object "pctable".

pctable <- PCoA(taxonomic.table = 'organised_genus_table.txt',
                meta = 'meta.txt', 
                group = 'Treatment', 
                background.variable = 'Richness', 
                readcount.cutoff = 2000,
                pdf = T, keep.results = T)
	

## End(Not run)

katrikorpela/mare documentation built on July 17, 2022, 2:49 a.m.