controlForConfounders: Autoencoder function to correct for confounders.

In gagneurlab/OUTRIDER: OUTRIDER - OUTlier in RNA-Seq fInDER

Autoencoder function to correct for confounders.

Description

This is the wrapper function for the autoencoder implementation. It can be used to call the standard R implementation or the experimental Python implementation.

Usage

controlForConfounders(
  ods,
  q,
  implementation = c("autoencoder", "pca"),
  BPPARAM = bpparam(),
  ...
)

Arguments

ods	An OutriderDataSet object
q	The encoding dimensions
implementation	"autoencoder", the default, will use the autoencoder implementation. Also 'pca' and 'peer' can be used to control for confounding effects
BPPARAM	A BiocParallelParam instance to be used for parallel computing.
...	Further arguments passed on to the specific implementation method.

Value

An ods object including the control factors

Examples

ods <- makeExampleOutriderDataSet()
implementation <- 'autoencoder'

ods <- estimateSizeFactors(ods)
ods <- controlForConfounders(ods, implementation=implementation)

plotCountCorHeatmap(ods, normalized=FALSE)
plotCountCorHeatmap(ods, normalized=TRUE)

gagneurlab/OUTRIDER documentation built on April 29, 2024, 2:22 a.m.