Skip to contents

Prepare class for contrastive PCA

contrastive_pca_processing.Rd

This function will prepare the BulkCoExp for subsequent usage of the contrastive PCA functions. This is based on the work of Abid, et al.

Usage

contrastive_pca_processing(
  object,
  background_matrix,
  scale = FALSE,
  .verbose = TRUE
)

Arguments

object

The underlying class, see BulkCoExp().

background_matrix

Numeric matrix. The background matrix you wish to remove. You should apply any data transformation to this matrix, too!

scale

Boolean. Shall the data be scaled. Defaults to FALSE.

.verbose

Boolean. Controls verbosity of the function.

Value

BulkCoExp with the needed data for contrastive PCA in the properties of the class.

References

Abid, et al., Nature Communications, 2018

Examples

# covariance matrices of the target and the background
cpca_data <- synthetic_c_pca_data()
target <- t(cpca_data$target)
background <- t(cpca_data$background)
meta <- data.table::data.table(sample_id = rownames(target))
obj <- BulkCoExp(target, meta)
obj <- preprocess_bulk_coexp(obj, .verbose = FALSE)
obj <- contrastive_pca_processing(obj, background, .verbose = FALSE)
dim(obj@processed_data$target_covar)
#> [1] 30 30