Learn R Programming

TreeTools (version 1.16.1)

LongBranch: Identify taxa with long branches

Description

The long branch (LB) score Struck2014TreeTools measures the deviation of the average pairwise patristic distance of a leaf from all other leaves in a tree, relative to the average leaf-to-leaf distance.

Usage

LongBranch(tree)

Value

LongBranch() returns a vector giving the long branch score for each leaf in tree, or a list of such vectors if tree is a list. Results are given as raw deviations, without multiplying by 100 as proposed by Struck2014;textualTreeTools.

Arguments

tree

A tree of class phylo, or a list of trees of class list or multiPhylo.

Details

Struck2014;textualTreeTools proposes the standard deviation of LB scores as a measure of heterogeneity that can be compared between trees; and the upper quartile of LB scores as "a representative value for the taxa with the longest branches".

See Also

Other tree properties: ConsensusWithout(), MatchEdges(), NSplits(), NTip(), NodeNumbers(), PathLengths(), SplitsInBinaryTree(), TipLabels(), TreeIsRooted(), Treeness()

Examples

Run this code
tree <- BalancedTree(8, lengths = c(rep(2, 4), 5:7, rep(2, 4), rep(1, 3)))
lb <- LongBranch(tree)
tree$tip.label <- paste(tree$tip.label, signif(lb, 3), sep = ": ")
plot(tree, tip.col = SupportColour((1 - lb) / 2), font = 2)

# Standard deviation of LB scores allows comparison with other trees
sd(lb)
evenLengths <- BalancedTree(8, lengths = jitter(rep(1, 14)))
sd(LongBranch(evenLengths))

# Upper quartile identifies taxa with longest branches
threshold <- quantile(lb, 0.75)
tree$tip.label[lb > threshold]

Run the code above in your browser using DataLab