marginaliseEqual: Computes marginal likelihoods that two replicate groups are...

In baySeq: Empirical Bayesian analysis of patterns of differential expression in count data

Description

In cases where multiple models are simultaneously evaluated in the 'getLikelihoods' function, the posterior likelihoods for each model in which two conditions are equivalent can be summed to give the marginal likelihood of equivalence for all biomolecular events (i.e., data rows).

Usage

marginaliseEqual(cD, r1, r2)

Arguments

cD	A countData object with evaluated posterior likelihoods in the '@posteriors' slot.
r1	A defined group name to identify in the '@groups' slot of the countData object 'cD'.
r2	A defined group name to identify in the '@groups' slot of the countData object 'cD'.

Value

A vector of marginal posterior likelihoods defining the probability that the two group identifiers are equal for each row of the data.

Author(s)

Thomas J. Hardcastle

See Also

allModels marginalisePairwise

Examples

# load test data
data(simData)

# Create a {countData} object from test data, supposing that there are
# multiple experimental groups present.

replicates <- c("simA", "simA", "simB", "simC", "simC", "simD", "simE", "simE", "simF", "simG")
CD <- new("countData", data = simData, replicates = replicates)
CD <- allModels(CD)

# The total number of models generated is high.
length(CD@groups)

# Priors and likelihoods acquired through standard means.

## Not run: CD <- getPriors(CD, cl = cl)
## Not run: CD <- getLikelihoods(CD, cl = cl)

# Marginal likelihood that 'simA' and 'simD' replicate groups are equal
# for each row of the data.

## Not run: marginaliseEqual(CD, "simA", "simD")

baySeq documentation built on Nov. 8, 2020, 5:43 p.m.