metaMix (version 0.1)

parallel.temper.nucl: Parallel Tempering MCMC

Description

Parallel Tempering MCMC

Performs Parallel Tempering MCMC to explore the species state space. Two types of moves are implemented: a mutation step (within chain) and an exchange step (between neighboring chains). If working with BLASTx data, use parallel.temper().

parallel.temper.nucl.explicit is the same function as parallel.temper.nucl with a more involved syntax.

Usage

parallel.temper.nucl(step2, readSupport = 10, noChains = 12, seed = 1,
  median.genome.length = 284332)
parallel.temper.nucl.explicit(readSupport = 10, noChains = 12,
  pij.sparse.mat, read.weights, ordered.species, gen.prob.unknown, outDir,
  seed = 1, median.genome.length = 284332)

Arguments

step2

list. The output from reduce.space(). Alternatively, it can be a character string containing the path name of the ".RData" file where step2 list was saved.

readSupport

The number of reads the user requires in order to believe in the presence of the species. It is used to compute the penalty factor. The default value is 10. We compute the logarithmic penalty value as the log-likelihood difference between two models: one

noChains

The number of parallel chains to run. The default value is 12.

seed

Optional argument that sets the random seed (default is 1) to make results reproducible.

median.genome.length

To use in the penalty computation.

pij.sparse.mat

sparse matrix of generative probabilities, see value of ?reduce.space.

read.weights

see ?reduce.space.

ordered.species

see ?reduce.space.

gen.prob.unknown

see ?reduce.space.

outDir

see ?reduce.space.

Value

• step3: A list with two elements. The first one (result) is a list that records MCMC information from each parallel chain. The second one (duration) records how much time the MCMC exploration took.

See Also

parallel.temper This function should be used when working with BLASTx data.