rand_tip: Tree expansion function.

In iramosgutierrez/randtip: A generalized framework to expand incomplete phylogenies with taxonomic information

Tree expansion function.

Description

Expand a phylogeny binding PUTs to a backbone tree.

Usage

rand_tip(
  input,
  tree,
  rand.type = "random",
  polyphyly.scheme = "largest",
  use.paraphyletic = TRUE,
  use.singleton = TRUE,
  use.stem = FALSE,
  respect.mono = TRUE,
  respect.para = TRUE,
  clump.puts = TRUE,
  prob = TRUE,
  prune = TRUE,
  forceultrametric = FALSE,
  verbose = TRUE
)

Arguments

input	An 'input' data frame obtained with info2input function.
tree	A backbone tree.
rand.type	For all PUTs not specified individually in 'input', which randomization type ("random" or "polytomy") must be carried out. Default value is "random".
polyphyly.scheme	For all PUTs not specified individually in 'input', which polyphyly scheme ("largest", "complete" or "frequentist") must be used. Default value is "largest".
use.paraphyletic	For all PUTs not specified individually in 'input', whether or not should paraphyletic clades be taken into account or not. Default value is TRUE.
use.singleton	For all PUTs not specified individually in 'input', should or not singleton MDCCs be considered for binding as a sister species, or contrarily binding should be performed anywhere below the parent node. Default value is TRUE.
use.stem	For all PUTs not specified individually in 'input', whether or not should the stem branch be considered as candidate for binding. Default value is FALSE.
respect.mono	For all PUTs not specified individually in 'input', whether or not monophyletic groups should be respected when binding a PUT. Default value is TRUE.
respect.para	For all PUTs not specified individually in 'input', whether or not paraphyletic groups should be respected when binding a PUT. Default value is TRUE.
clump.puts	For all PUTs not specified individually in 'input', whether or not co-ranked PUTs should be clumped together in the phylogeny in case their taxonomic group is missing in the tree. Will also clump conspecific PUTs. Default value is TRUE.
prob	For all PUTs not specified individually in 'input', whether or not branch selection probability must be proportional to branch length or equiprobable. Default value is TRUE.
prune	Whether or not the newly expanded tree will include only the species in the user's list. Default value is TRUE.
forceultrametric	Whether or not the backbone tree will be forced to be ultrametric, only in case it is not. Default value is FALSE.
verbose	Whether or not to print information about the flow of the function. Default value is TRUE.

Details

The optimum binding procedure might be different in each case, given the phyletic nature of the MDCC to which the PUT should be bound to, and different reasons might drive the user to choose one or another.

• polyphyly.scheme The resulting outputs may be very different depending on the selected parameter. If 'complete' procedure is selected, the binding will be performed as a monophyletic group using the MRCA of all the known PPCR species. The 'largest' procedure, otherwise, will use as binding group the largest monophyletic group formed by PPCR species. Ultimately, the 'frequentist' option will split the group in monophyletic and singleton chunks and for every PUT to be bound to the MDCC, one of the chunks will be selected weighting the probability by the number of original tips.

• use.paraphyletic This parameter should be used as TRUE in case we are certain of the paraphyletic nature of one group, but disregarded if the "intruder" group within the otherwise monophyletic group is possibly due to an incorrect placement in the creation of the phylogeny.

• forceultrametric It is important to note that forcing phylogenies to be ultrametric in this way should not be taken as a formal statistical approach for inferring an ultrametric tree but a method to be deployed whenever a genuinely ultrametric phylogeny read from file fails due to issues related to numerical precision (Revell, 2012). Thus, we strongly recommend the user to visually explore phylogenetic trees that fail the ultrametricity test of check.info before assuming the failure is due to numerical precision of computer machinery.

Value

An expanded phylogeny.

Author(s)

Ignacio Ramos-Gutierrez, Rafael Molina-Venegas, Herlander Lima

Examples

 catspecies <- c("Lynx_lynx", "Panthera_uncia",
"Panthera_onca", "Felis_catus", "Puma_concolor",
"Lynx_canadensis", "Panthera_tigris", "Panthera_leo",
"Felis_silvestris")

cats.info <- build_info(species=catspecies, tree= cats,
     find.ranks=TRUE, db="ncbi", mode="backbone")

cats.input <- info2input(info=cats.info, tree=cats)

expanded.cats <- rand_tip(input=cats.input,
 tree=cats, rand.type = "random",
  forceultrametric = FALSE)

iramosgutierrez/randtip documentation built on March 19, 2024, 6:35 p.m.