tess.likelihood.ebdstp: tess.likelihood.ebdstp

In hoehna/TESS: Diversification Rate Estimation and Fast Simulation of Reconstructed Phylogenetic Trees under Tree-Wide Time-Heterogeneous Birth-Death Processes Including Mass-Extinction Events

Description

Computation of the likelihood for a given tree under an episodic fossilized-birth-death model (i.e. piecewise constant rates).

Usage

tess.likelihood.ebdstp(
  nodes,
  lambda,
  mu,
  phi,
  r,
  samplingProbabilityAtPresent,
  rateChangeTimesLambda = c(),
  rateChangeTimesMu = c(),
  rateChangeTimesPhi = c(),
  rateChangeTimesR = c(),
  massDeathTimes = c(),
  massDeathProbabilities = c(),
  burstBirthTimes = c(),
  burstBirthProbabilities = c(),
  eventSamplingTimes = c(),
  eventSamplingProbabilities = c(),
  samplingStrategyAtPresent = "uniform",
  MRCA = TRUE,
  CONDITION = "survival",
  log = TRUE
)

Arguments

nodes	node times from tess.branching.times
lambda	birth (speciation or infection) rates
mu	death (extinction or becoming non-infectious without treatment) rates
phi	serial sampling (fossilization) rates
r	treatment probability, Pr(death | sample) (does not apply to samples take at time 0)
samplingProbabilityAtPresent	probability of uniform sampling at present
rateChangeTimesLambda	times at which birth rates change
rateChangeTimesMu	times at which death rates change
rateChangeTimesPhi	times at which serial sampling rates change
rateChangeTimesR	times at which treatment probabilities change
massDeathTimes	time at which mass-deaths (mass-extinctions) happen
massDeathProbabilities	probability of a lineage dying in a mass-death event
burstBirthTimes
burstBirthProbabilities
eventSamplingTimes	time at which every lineage in the tree may be sampled
eventSamplingProbabilities	probability of a lineage being sampled at an event sampling time
samplingStrategyAtPresent	Which strategy was used to obtain the samples (taxa). Options are: uniform|diversified|age
MRCA	does the tree start at the mrca?
CONDITION	do we condition the process on nothing|survival|taxa?
log	likelhood in log-scale?

Value

probability of the speciation times

Examples

data(conifers)
  
nodes <- tess.branching.times(conifers)

lambda <- c(0.2, 0.1, 0.3)
mu <- c(0.1, 0.05, 0.25)
phi <- c(0.1, 0.2, 0.05)

changetimes <- c(100, 200)

tess.likelihood.ebdstp(nodes,
                       lambda = lambda, 
                       mu = mu,
                       phi = phi,
                       r = 0.0,
                       rateChangeTimesLambda = changetimes,
                       rateChangeTimesMu = changetimes,
                       rateChangeTimesPhi = changetimes,
                       samplingProbability = 1.0
)

hoehna/TESS documentation built on Feb. 3, 2022, 5:59 a.m.