minBranchLength {paleotree}

Scales Edge Lengths of a Phylogeny to a Minimum Branch Length

Description

Rescales a tree with edge lengths so that all edge lengths are at least some minimum branch length (sometimes abbreviated as "MBL" or "mbl"). Edge lengths are transformed so they are greater than or equal to the input minimum branch length, by subtracting edge length from more root-ward edges and added to later branches. This may or may not change the age of the root divergence, depending on the distribution of short branch lengths close to the root.

Usage

minBranchLength(tree, mbl, modifyRootAge = TRUE)

Arguments

tree

A phylogeny with edge lengths of class phylo.

mbl

The minimum branch length

modifyRootAge

If TRUE (the default), the input tree is checked for a root age given as $root.time and if present it is checked and fixed for any possible movement backwards due to short branches close to the root node.

Details

This function was formally an internal segment in timePaleoPhy, and now is called by timePaleoPhy instead, allowing users to apply minBranchLength to trees that already have edge lengths.

Value

A phylogeny with edge lengths of class phylo.

Author(s)

David W. Bapst

See Also

This function was originally an internal piece of timePaleoPhy, which implements the minimum branch length time-scaling method along with others, which may be what you're looking for (instead of this miscellaneous function).

Examples

#simulation with an example non-ultrametric tree

tree <- rtree(20)
# randomly replace edges with ZLBs
   # similar to multi2di output
tree <- degradeTree(tree,0.3,
   leave.zlb = TRUE) 	

tree2 <- minBranchLength(tree,0.1)

layout(1:2)

plot(tree)
axisPhylo()
plot(tree2)
axisPhylo()

layout(1)


#now let's try it with an ultrametric case

# get a random tree
tree <- rtree(30)
# randomly replace edges with ZLBs
   # similar to multi2di output
tree <- degradeTree(tree,0.5,leave.zlb = TRUE) 
# now randomly resolve	
tree <- di2multi(tree)
# give branch lengths so its ultrametric
tree <- compute.brlen(tree)

# and we have an ultrametric tree with polytomies, yay!
plot(tree) 

# now randomly resolve
tree2 <- multi2di(tree)
# get new branch lengths as would with real data
tree2 <- minBranchLength(tree2,0.1)

layout(1:2)
plot(tree,show.tip.label = FALSE)
axisPhylo()
plot(tree2,show.tip.label = FALSE)
axisPhylo()

layout(1)

# check that root ages aren't being left unmodified
   # create a tree with lots of ZBLs at the root
x <- stree(10)
x$edge.length <- runif(Nedge(x))
x <- multi2di(x)
# give it a root age
x$root.time <- max(node.depth.edgelength(x))

z <- minBranchLength(tree = x, mbl = 1)
plot(z)