tess.process.output: tess.process.output: Summarizing the output of a...

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

tess.process.output: Summarizing the output of a diversification rate estimation including mass-extinction events. See the tess.analysis function for more information on how such output is generated and the tess.plot.output how the output can be visualized. Also have a look at the vignette for more in detail description and examples.

Description

tess.process.output summarizes the output generated by a tess.analysis(...) run.

Usage

tess.process.output(dir,
                    tree=NULL,
                    numExpectedRateChanges=2,
                    numExpectedMassExtinctions=2,
                    burnin=0.25,
                    numIntervals=100,
                    criticalBayesFactors=c(2,6,10))

Arguments

dir	The directory from which the CoMET output will be read.
tree	The tree analyzed with CoMET in phylo format. By default, looks for a tree in the target directory.
numExpectedRateChanges	The number of expected diversification-rate changes.
numExpectedMassExtinctions	The number of expected mass-extinction events.
burnin	The fraction of samples that will be discarded as burnin.
numIntervals	The number of discrete intervals in which to break the tree.
criticalBayesFactors	The Bayes factor thresholds to use to assess significance of events.

Details

The output of a CoMET analysis is stored in a directory with different files containing the MCMC samples from the posterior distribution. For example, the tess.analysis function stores the times and survival probabilities of the mass-extinction events in a file. This function, converts the output by counting the number of events that fall into a given time-bin. This pre-processing of the output simplifies the plotting.

Value

This function returns a list with the following elements:

posterior	An object of class 'mcmc' that contains the trace of the model's posterior probability.
numSpeciationCategories	An object of class 'mcmc' that contains samples from the posterior distribution of the number of speciation categories (minimum 1, since this includes the initial speciation rate).
numExtinctionCategories	An object of class 'mcmc' that contains samples from the posterior distribution of the number of extinction categories (minimum 1, since this includes the initial extinction rate).
numMassExtinctions	An object of class 'mcmc' that contains samples from the posterior distribution of the number of mass-extinction events.
speciation rates	An object of class 'mcmc' that contains speciation rates sampled from the posterior distribution for each of numIntervals discrete time intervals. Rows correspond to samples from the posterior distribution, columns correspond to intervals.
speciation change times	An object of class 'mcmc' that contains speciation-rate-change events sampled from the posterior distribution for each of numIntervals discrete time intervals. A value of 1 indicates an event was contained in the interval, 0 that no event was contained in the interval. Rows correspond to samples from the posterior distribution, columns correspond to intervals.
speciation Bayes factors	A vector of class 'numeric' that contains the Bayes factor support for there being a speciation-rate-change event for each of numIntervals discrete time intervals. The ith element corresponds to the Bayes factor support for an event in the ith interval.
speciationRateChangeCriticalPosteriorProbabilities	A vector of posterior probabilities that correspond to critical Bayes factor thresholds (specified by the argument criticalBayesFactors). Element i is the posterior probability of a speciation-rate-change event in an interval needed to produce Bayes factor support of criticalBayesFactors[i].
extinction rates	An object of class 'mcmc' that contains extinction rates sampled from the posterior distribution for each of numIntervals discrete time intervals. Rows correspond to samples from the posterior distribution, columns correspond to intervals.
extinction change times	An object of class 'mcmc' that contains extinction-rate-change events sampled from the posterior distribution for each of numIntervals discrete time intervals. A value of 1 indicates an event was contained in the interval, 0 that no event was contained in the interval. Rows correspond to samples from the posterior distribution, columns correspond to intervals.
extinction Bayes factors	A vector of class 'numeric' that contains the Bayes factor support for there being a extinction-rate-change event for each of numIntervals discrete time intervals. The ith element corresponds to the Bayes factor support for an event in the ith interval.
extinctionRateChangeCriticalPosteriorProbabilities	A vector of posterior probabilities that correspond to critical Bayes factor thresholds (specified by the argument criticalBayesFactors). Element i is the posterior probability of a extinction-rate-change event in an interval needed to produce Bayes factor support of criticalBayesFactors[i].
net-diversification rates	An object of class 'mcmc' that contains net-diversification (speciation - extinction) rates sampled from the posterior distribution for each of numIntervals discrete time intervals. Rows correspond to samples from the posterior distribution, columns correspond to intervals.
relative-extinction rates	An object of class 'mcmc' that contains relative-extinction (extinction / speciation) rates sampled from the posterior distribution for each of numIntervals discrete time intervals. Rows correspond to samples from the posterior distribution, columns correspond to intervals.
mass extinction times	An object of class 'mcmc' that contains mass-extinction events sampled from the posterior distribution for each of numIntervals discrete time intervals. A value of 1 indicates an event was contained in the interval, 0 that no event was contained in the interval. Rows correspond to samples from the posterior distribution, columns correspond to intervals.
mass extinction Bayes factors	A vector of class 'numeric' that contains the Bayes factor support for there being a mass-extinction event for each of numIntervals discrete time intervals. The ith element corresponds to the Bayes factor support for an event in the ith interval.
massExtinctionCriticalPosteriorProbabilities	A vector of posterior probabilities that correspond to critical Bayes factor thresholds (specified by the argument criticalBayesFactors). Element i is the posterior probability of a mass-extinction event in an interval needed to produce Bayes factor support of criticalBayesFactors[i].
criticalBayesFactors	The critical Bayes factor values used for the Bayes factor tests (default 2 ln BF = {2,6,10}).
tree	The tree analyzed with CoMET (just in case).
intervals	The discrete intervals used to compute the interval-specific parameters.

Author(s)

Michael R. May

Examples

# Load the data, compute the sampling fraction rho
data(conifers)
totalConiferSpecies <- 630
sampledConiferSpecies <- conifers$Nnode+1
rho <- sampledConiferSpecies / totalConiferSpecies

# Run a tess analysis
tess.analysis(tree = conifers,
              initialSpeciationRate=c(1.0),
              initialExtinctionRate=c(0.5),
              empiricalHyperPriors = FALSE,
              numExpectedRateChanges = 2,
              numExpectedMassExtinctions = 2,
              samplingProbability = rho,
              MAX_ITERATIONS = 200,
              BURNIN=100)

# Process the output
coniferOutput <- tess.process.output(dir=getwd(),
                                     numExpectedRateChanges=2,
                                     numExpectedMassExtinctions=2)

# Plot the output
tess.plot.output(coniferOutput)

TESS documentation built on April 25, 2022, 5:07 p.m.