Get module preservation statistics

Tests whether a set of modules defined in the reference dataset are preserved in the test dataset. Provides a convenient wrapper around modulePreservation for ModularExperiment and SummarizedExperiment objects.

Usage

modulePreservation(
  reference_dataset,
  test_dataset,
  reference_assay_name = "normal",
  test_assay_name = "normal",
  module_assignments = NULL,
  greyName = "module_0",
  goldName = "random",
  networkType = "signed",
  corFnc = "cor",
  savePermutedStatistics = FALSE,
  ...
)

Arguments

reference_dataset

The dataset that was used to define the modules. Must be a data.frame or matrix with features as rows and samples as columns, or a ModularExperiment or SummarizedExperiment object.

test_dataset

The dataset that will be used to test for module preservation. Must be a data.frame or matrix with features as rows and samples as columns, or a SummarizedExperiment object. The features of test_dataset should be the same as reference_dataset and in the same order.

reference_assay_name

If the reference dataset is a ModularExperiment or SummarizedExperiment object, this argument specifies which assay slot was used to define the modules.

test_assay_name

If the reference dataset is a ModularExperiment or SummarizedExperiment object, this argument specifies which assay slot is to be used in preservation tests.

module_assignments

If the reference dataset is not a ModularExperiment object, this argument is necessary to specify the module assignments.

greyName

The name of the "module" of unassigned genes. Usually "module_0" (ReducedExperiment default) or "grey" (WGCNA default). See modulePreservation.

goldName

The name to be used for the "gold" module (which is made up of a random sample of all network genes). See modulePreservation.

networkType

A string referring to the type of WGCNA network used for the reference and test datasets. One of"unsigned", "signed" or "signed hybrid". See adjacency. See modulePreservation.

corFnc

A string referring to the function to be used to calculate correlation. One of "cor" or "bicor". See modulePreservation.

savePermutedStatistics

If TRUE, saves the permutation statistics as a .RData file. See modulePreservation.

...

Additional arguments to be passed to modulePreservation.

Value

A data.frame containing preservation statistics, as described by modulePreservation.

Author

Jack Gisby

Examples

# Get random ModularExperiments with rnorm, with 100 rows (features),
# 20 columns (observations) and 5/10 modules
me_1 <- ReducedExperiment:::.createRandomisedModularExperiment(100, 20, 5)
me_2 <- ReducedExperiment:::.createRandomisedModularExperiment(100, 20, 10)

# Test module preservation (test modules from dataset 1 in dataset 2)
mp <- modulePreservation(me_1, me_2, verbose = 0, nPermutations = 3)