phylo.to.sim {paleobuddy}

Converting a phylogeny in a paleobuddy object

Description

Generates a sim object using a phylo object and some additional information (depending on other arguments). It is the inverse of the make.phylo function. Input is (1) a phylogeny, following an evolutionary Hennigian (sensu Ezard et al 2011) format (i.e., a fully bifurcating phylogeny), (2) information on the "mother lineage" of each tip in the phylogeny, (3) the status ("extant" or "extinct") of each lineage, (4) the stem age (or age of origination of the clade), and (5) the stem length (or time interval between the stem age and the first speciation event). The user can also choose if the event dating should be done from root to tips or from tips-to-root. The function returns a sim object (see ?sim). The function does not accept more than one species having NA as parent (which is interpreted as if there were no single common ancestor in the phylogeny). In that case, use find.lineages first.

Usage

phylo.to.sim(
  phy,
  mothers,
  extant,
  dateFromPresent = TRUE,
  stemAge = NULL,
  stemLength = NULL
)

Arguments

phy

A phylo object, which may contain only extant or extant and extinct lineages.

mothers

Vector containing the mother of each tip in the phylogeny. First species' mother should be NA. See details below.

extant

Logical vector indicating which lineages are extant and extinct.

dateFromPresent

Logical vector indicating if speciation/extinction events should be dated from present-to-root (TRUE, default value) or from root-to-present. As it is impossible to date "from present" without a living lineage, it is internally set to FALSE and prints a message in the prompt if there are no extant species.

stemAge

Numeric vetor indicating the age, in absolute geological time (million years ago), when the first lineage of the clade originated. It is not needed when dateFromPresent is TRUE and stemLength is provided, or when phy has a root.edge. This argument is required if dateFromPresent is FALSE.

stemLength

Numeric vector indicating the time difference between the stemAge and the first speciation event of the group. This argument is required if dateFromPresent is FALSE, but users have no need to assign values in this parameter if phy has a $root.edge, which is taken by the function as the stemLength value.

Details

See Details below for more information on each argument.

Mothers:

The function needs the indication of a mother lineage for every tip in the phylogeny but one (which is interpreted as the first known lineage in the clade, and should have NA as the mother). This assignment might be straightforward for simulations (as in the examples section below), but is a non-trivial task for empirical phylogenies. As there are many ways to assign impossible combinations of motherthood, the function does not return any specific error message if the provided motherhood does not map to possible lineages given the phylogeny. Instead, the function tends to crash when an "impossible" motherhood is assigned, but this is not guaranteed to happen because the set of "impossible" ways to assign motherhood is vast, and therefore has not allowed for a test of every possibility. If the function crashes when all lineages have reasonable motherhood, users should submit an issue report at https://github.com/brpetrucci/paleobuddy/issues.

Dating:

Phylogenies store the relative distances between speciation (and possibly extinction) times of each lineage. However, to get absolute times for those events (which are required to construct the output of this function), users should provide a moment in absolute geological time to position the phylogeny. This could be (1) the present, which is used as reference in the case at least one lineage in the phylogeny is extant (i.e., default behavior of the function), or (2) some time in the past, which is the stemAge parameter. Those two possible dating methods are used by setting dateFromPresent to TRUE or FALSE. If users have extant lineages in their phylogeny but do not have a reasonable value for stemAge, they are encouraged to use present-to-root dating (dateFromPresent = TRUE), as in that case deviations in the value of stemLength will only affect the speciation time of the first lineage of the clade. In other words, when dateFromPresent is set to FALSE, user error in stemAge or stemLength will bias the absolute (but not the relative) dating of all nodes in the phylogeny.

Value

A sim object. For details, see ?sim. Items in the object follow their tip assignment in the phylogeny.

Author(s)

Matheus Januario.

References

Ezard, T. H., Pearson, P. N., Aze, T., & Purvis, A. (2012). The meaning of birth and death (in macroevolutionary birth-death models). Biology letters, 8(1), 139-142.

Examples

# to check the usage of the function, let us make sure it transforms a 
# phylogeny generated with make.phylo back into the original simulation

### 
# birth-death process

# set seed
set.seed(1)

# run simulation
sim <- bd.sim(1, lambda = 0.3, mu = 0.1, tMax = 10, nFinal = c(10, Inf))

# convert birth-death into phylo
phy <- make.phylo(sim)

# convert phylo into a sim object again
res <- phylo.to.sim(phy = phy, extant = sim$EXTANT, mothers = sim$PAR)

# test if simulation and converted object are the same
all.equal(sim, res)

### 
# birth-death process with extinct lineages:
# set seed
set.seed(1)

# run simulation
sim <- bd.sim(1, lambda = 0.1, mu = 0.3, tMax = 10, nFinal = c(2, 4))

# convert birth-death into phylo
phy <- make.phylo(sim)

# convert phylo into a sim object again
res <- phylo.to.sim(phy = phy, extant = sim$EXTANT, mothers = sim$PAR, stemAge = max(sim$TS))

# test if simulation and converted object are the same
all.equal(sim, res)

###
# pure birth process

# set seed
set.seed(1)

# run simulation
sim <- bd.sim(1, lambda = 0.2, mu = 0, tMax = 10, nFinal = c(10, Inf))

# convert birth-death into phylo
phy <- make.phylo(sim)

# convert phylo into birth-death again
# note we can supply optional arguments, see description above
res <- phylo.to.sim(phy = phy, extant = sim$EXTANT, mothers = sim$PAR, 
                stemAge = 10, stemLength = (10 - sim$TS[2]))

# testing if simulation and converted object are the same
all.equal(sim, res)