Build a novel phylogeny from existing data
congeneric.impute sequentially add species to a phylogeny to
form an _imputed_ bifurcating tree. Makes use of a result from
Steel & Mooers (2010) that gives the expected branch-length under a
Yule model whether the rate of diversification has been
estimated. The intention of this is to approximate the method by
which phylogenetic structure is sampled from the prior in BEAST;
i.e., to approximate the standard Kuhn et al. (2011) method for
imputing a phylogeny. When using congeneric.impute you
should (1) repeat your analyses across many (if in doubt,
thousands) of separate runs (see Kuhn et al. 2011) and (2) check
for yourself that your trees are unbiased for your purpose - I make
no guarantee this is appropriate, and in many cases I think it
would not be. See also 'notes' below.
congeneric.merge(tree, species, split = "_", ...)bind.replace(backbone, donor, replacing.tip.label, donor.length = NA)
congeneric.impute(tree, species, split = "_", max.iter = 1000, ...)
phylo phylogeny
phylogeny (phylo)
phylo phylogeny to have those
species inserted into it
vector of species names to be bound into the tree if missing from it
the character that splits genus and species names in
your phylogeny. Default is _, i.e. Quercus_robur.
ignored
backbone phylogeny (phylo)
into which the donor is to be bound
phylogeny (phylo) to bound
into the backbone phylogeny
the species in the donor phylogeny that's being replaced by the donor phylogeny
how deep the donor phylogeny should be cut into the backbone phylogeny. If NA (default), then the bladj algorithm is followed (or, in plain English, it's put half-way along the branch)
Sometimes the random draw for the new branch length
to be added will be too large to allow it to be added to the
tree. In such cases, congeneric.imput will randomly draw
another branch length, and it will repeat this process
max.iter times. See 'notes' for more on this.
Will Pearse
Will Pearse
congeneric.merge Binds missing species into a
phylogeny by replacing all members of the clade it belongs to with
a polytomy. Assumes the tip.labels represent Latin
binomials, split by the split argument. This code was
originally shipped with phyloGenerator - this is the merge
method in that program.
bind.replace Binds a phylogeny (donor) into a
bigger phylogeny ('backbone'); useful if you're building a
phylogeny a la Phylomatic. A version of this R code was shipped
with phyloGenerator (Pearse & Purvis 2013). This is really an
internal function for congeneric.merge, but hopefully it's
of some use to you!
Pearse W.D. & Purvis A. phyloGenerator: an automated phylogeny generation tool for ecologists. Methods in Ecology and Evolution 4(7): 692--698.
Steel, M., & Mooers, A. (2010). The expected length of pendant and interior edges of a Yule tree. Applied Mathematics Letters, 23(11), 1315-1319.
Kuhn, T. S., Mooers, A. O., & Thomas, G. H. (2011). A simple polytomy resolver for dated phylogenies. Methods in Ecology and Evolution, 2(5), 427-436.
tree <- read.tree(text="((a_a:1,b_b:1):1, c_c:2):1;")
tree <- congeneric.merge(tree, c("a_nother", "a_gain", "b_sharp"))
tree <- read.tree(text="((a_a:1,b_b:1):1, c_c:2):1;")
tree <- congeneric.impute(tree, c("a_nother", "a_gain", "b_sharp"))
Run the code above in your browser using DataLab