shrinkBins: Employ an eBayes shrinkage approach for bin-level estimates...

In biobenkj/compartmentalizer: Higher-order chromatin domain inference in single samples from methylation arrays and single cells from scRNA-seq and scATAC-seq

Employ an eBayes shrinkage approach for bin-level estimates for A/B inference

Description

shrinkBins returns shrunken bin-level estimates

Usage

shrinkBins(
  x,
  original.x,
  prior.means = NULL,
  chr = NULL,
  res = 1e+06,
  targets = NULL,
  jse = TRUE,
  assay = c("rna", "atac", "array"),
  genome = c("hg19", "hg38", "mm9", "mm10")
)

Arguments

x	Input SummarizedExperiment object
original.x	Full sample set SummarizedExperiment object
prior.means	The means of the bin-level prior distribution
chr	The chromosome to operate on
res	Resolution to perform the binning
targets	The column/sample/cell names to shrink towards
jse	Whether to use a James-Stein estimator (default is TRUE)
assay	What assay type this is ("rna", "atac", "array")
genome	What genome are we working with ("hg19", "hg38", "mm9", "mm10")

Details

This function computes shrunken bin-level estimates using a James-Stein estimator, reformulated as an eBayes procedure

Value

A list object to pass to getCorMatrix

Examples

data("k562_scrna_chr14", package = "compartmap")
shrunken.bin.scrna <- shrinkBins(x = k562_scrna_chr14,
                                 original.x = k562_scrna_chr14,
                                 chr = "chr14", assay = "rna")

biobenkj/compartmentalizer documentation built on Oct. 13, 2023, 5:26 p.m.