R: Plot a Phylogeny and Traits

trait.plot {diversitree}

R Documentation

Plot a Phylogeny and Traits

Description

Plot a phylogeny and label the tips with binary traits. This function is experimental, and may change soon. Currently it is limited to radial trees and binary traits.

Usage

trait.plot(tree, dat, cols, lab=names(cols), str=0:1, class=NULL,
           type="f", w=1/50, legend=length(cols) > 1, cex.lab=.5,
           font.lab=3, cex.legend=.75, margin=1/4,
           check=TRUE, quiet=FALSE)

Arguments

`tree`	Phylogenetic tree, in ape format.
`dat`	A `data.frame` of trait values. The row names must be the same names as the tree (`tree$tip.label`), and each column contains the states (0, 1, or `NA`). The column names must give the trait names.
`cols`	A list, each element of which is a vector of length 2 being the colours for state 0 and state 1 for each trait. The list must be named so that all names appear in `names(dat)`. Traits will be plotted in the order given by `cols`.
`lab`	Alternative names for the legend (perhaps longer or more informative). Must be in the same order as `cols`.

`str`	Strings used for the states in the legend.
`class`	A vector along `phy$tip.label` giving a higher level classification (e.g., genus or family). No checking is done to ensure that such classifications are not polyphyletic.
`type`	Plot type (same as `type` in `?plot.phylo`). Currently only `f` (fan) is implemented.
`w`	Width of the trait plot, as a fraction of the tree depth.
`legend`	Logical: should a legend be plotted?
`cex.lab, font.lab`	Font size and type for the tip labels.
`cex.legend`	Font size for the legend.
`margin`	How much space, relative to the total tree depth, should be reserved when plotting a higher level classification.
`check`	When TRUE (by default), this will check that the classes specified by `class` are monophyletic. If not, classes will be concatenated and a warning raised.
`quiet`	When TRUE (FALSE by default), this suppresses the warning caused by `check=TRUE`.

Author(s)

Richard G. FitzJohn

Examples

These are the parameters: they are a single speciation and extinction rate, then 0->1 (trait A), 1->0 (A), 0->1 (B) and 1->0 (B).

colnames(musse.multitrait.translate(2, depth = 0))

## [1] "lambda0" "mu0"     "qA01.0"  "qA10.0"  "qB01.0"  "qB10.0"

Simulate a tree where trait A changes slowly and B changes rapidly.

set.seed(1)
phy <- tree.musse.multitrait(c(0.1, 0, 0.01, 0.01, 0.05, 0.05), n.trait = 2, 
    depth = 0, max.taxa = 100, x0 = c(0, 0))

Here is the matrix of tip states (each row is a species, each column is a trait).

head(phy$tip.state)

##     A B
## sp1 0 0
## sp2 0 1
## sp3 1 1
## sp4 0 1
## sp5 0 0
## sp6 0 0


trait.plot(phy, phy$tip.state, cols = list(A = c("pink", "red"), 
    B = c("lightblue", "blue")))

plot of chunk unnamed-chunk-3


nodes <- c("nd5", "nd4", "nd7", "nd11", "nd10", "nd8")
descendants <- function(node, phy) {
    ans <- node <- match(node, phy$node.label) + length(phy$tip.label)
    repeat {
        node <- phy$edge[phy$edge[, 1] %in% node, 2]
        if (length(node) > 0) 
            ans <- c(ans, node) else break
    }

    ans <- unlist(ans)
    ans[ans <= length(phy$tip.label)]
}

grp <- lapply(nodes, descendants, phy)
class <- rep(NA, 100)
for (i in seq_along(grp)) class[grp[[i]]] <- paste("group", LETTERS[i])

Now, 'class' is a vector along phy$tip.label indicating which of six groups each species belongs.

Plotting the phylogeny with these groups:

trait.plot(phy, phy$tip.state, cols = list(A = c("pink", "red"), 
    B = c("lightblue", "blue")), class = class, font = 1, cex.lab = 1, cex.legend = 1)

## Warning message: 'group.label.tip.rad' is deprecated.
## Use 'group.label.tip' instead.
## See help("Deprecated")

plot of chunk unnamed-chunk-4

[Package diversitree version 0.9-4 ]