computeWeightedDeviations

Computes weighted deviations in chromatin accessibility across sets of weights where the number of the annotation meaningful.

computeWeightedDeviations(object, weights, ...)


  # S4 method for SummarizedExperiment,SummarizedExperiment
computeWeightedDeviations(object,
  weights, background_peaks = getBackgroundPeaks(object),
  expectation = computeExpectations(object))

# S4 method for SummarizedExperiment,Matrix
computeWeightedDeviations(object,
  weights, background_peaks = getBackgroundPeaks(object),
  expectation = computeExpectations(object))

# S4 method for SummarizedExperiment,matrix
computeWeightedDeviations(object,
  weights, background_peaks = getBackgroundPeaks(object),
  expectation = computeExpectations(object))

Arguments

object

chromVARCounts object
weights

chromVARweights object
...

additional arguments
background_peaks

(optional) background peaks matrix computed using getBackgroundPeaks, computed internally with default paramaters if not provided
expectation

(optional) expectations computed using computeExpectations, computed automatically if not provided

Value

chromVARDeviations-class, which inherits from SummarizedExperiment, and has two assays: deviations and deviation scores.

Details

multiprocessing using bplapply

Methods (by class)

  • object = SummarizedExperiment,weights = SummarizedExperiment: / weights are SummarizedExperiment

  • object = SummarizedExperiment,weights = Matrix: object is SummarizedExperiment, weights are in a Matrix

  • object = SummarizedExperiment,weights = matrix: object is SummarizedExperiment, weights are in a numeric

Examples

# Register BiocParallel
BiocParallel::register(BiocParallel::SerialParam())
# Load very small example counts (already filtered)
data(mini_counts, package = "chromVAR")
# Load mini weighted counts
rdsA<-paste0(system.file('rds',package='gchromVAR'),'/mini_w.rds')
w_se <- readRDS(rdsA)

# Build weights from .bed file
files <- list.files(system.file('extdata',package='gchromVAR'),
         full.names = TRUE, pattern = "*.bed$")
data(mini_counts, package = "chromVAR")
uk_se <- importBedScore(SummarizedExperiment::rowRanges(mini_counts), files)

wdev <- computeWeightedDeviations(mini_counts, w_se)
ukdev <- computeWeightedDeviations(mini_counts, uk_se)