estimate_marginal_models: estimate the marginal likelihood of the relaxed and strict...

In thijsjanzen/nodeSub: Simulate DNA Alignments Using Node Substitutions

estimate the marginal likelihood of the relaxed and strict clock model for a provided alignment

Description

estimate_marginal_models estimates the marginal likelihood of both the strict and the relaxed clock model, given the JC69 substitution model, using the NS package in BEAST, made available via the babette R package. The NS package performs nested sampling, and uses an MCMC approach to estimate the marginal likelihood. Sampling is performed until convergence of the MCMC chain. Unfortunately, currently the babette package is unavailable on CRAN, requiring installation through GitHub to enjoy the full functionality of this function.

Usage

estimate_marginal_models(
  fasta_filename,
  use_yule_prior = FALSE,
  rng_seed = 42,
  sub_rate = 1,
  verbose = FALSE
)

Arguments

fasta_filename	file name of fasta file holding alignment for which the marginal likelihood is to be estimated
use_yule_prior	by default, a birth-death prior is used as tree prior, but if use_yule_prior is set to TRUE, a pure-birth prior will be used.
rng_seed	seed of pseudo-random number generator
sub_rate	substitution rate
verbose	boolean indicating if verbose intermediate output is to be generated

Value

data frame with marginal likelihoods and relative weights per clock model.

thijsjanzen/nodeSub documentation built on Nov. 15, 2023, 2:44 p.m.