imputeEmbedding: Impute values using StabMap joint embedding

In MarioniLab/StabMap: Stabilised mosaic single cell data integration using unshared features

Impute values using StabMap joint embedding

Description

Performs naive imputation of values from the list of mosaic data and joint embedding from StabMap.

Usage

imputeEmbedding(
  assay_list,
  embedding,
  reference = Reduce(union, lapply(assay_list, colnames)),
  query = Reduce(union, lapply(assay_list, colnames)),
  neighbours = 5,
  fun = mean
)

Arguments

assay_list	List of mosaic data from which to perform imputation.
embedding	Joint embedding from which to extract nearest neighbour relationships.
reference	Character vector of cell names to treat as reference cells.
query	Character vector of cell names to treat as query cells.
neighbours	Number of nearest neighbours to consider (default 5).
fun	function (default ‘mean') to aggregate nearest neighbours’ imputed values.

Value

List containing imputed values from each assay_list data matrix which contains reference cells.

Examples

set.seed(2021)
assay_list = mockMosaicData()
lapply(assay_list, dim)

# stabMap
out = stabMap(assay_list,
              ncomponentsReference = 20,
              ncomponentsSubset = 20)

# impute values
imp = imputeEmbedding(assay_list, out)

# inspect the imputed values
lapply(imp, dim)
imp[[1]][1:5,1:5]

MarioniLab/StabMap documentation built on Sept. 28, 2022, 2:28 a.m.