estimateDepthFactors: estimate library size correction factors
In YosefLab/MPRAnalyze: Statistical Analysis of MPRA data

Description Usage Arguments Value Note Examples

estimate library size correction factors

estimateDepthFactors(
  obj,
  lib.factor = NULL,
  which.lib = "both",
  depth.estimator = "uq"
)

`obj`	the MpraObject
`lib.factor`	the factor associating each sample to a library. Can be a factor or the name of a column in the object's colAnnot. If not provided, the data is assumed to have been generated from a single library, and constant library depth is set.
`which.lib`	which library to compute the depth factors for. Options are "both" (default), "dna" or "rna". If the DNA and RNA counts have different library factors, this function should be called twice: once with "dna" and once with "rna"
`depth.estimator`	a character indicating which depth estimation to use, or a function to perform the estimation. Currently supported values are "uq" for upper quantile of non-zero values (default), "rle" for RLE (uses geometric mean, and is therefore not recommended if libraries have 0 counts), or "totsum" for total sum. For a function input: function should take a numeric vector and return a single numeric, and preferably handle NA values. See examples.

the MpraObject with estimated values for sequencing depth factors

since in most MPRA experiments multiple barcodes exist within a single library, each column in the matrix is usually not a separate library. For this reason, it is recommended to supply this function with the appropriate partitioning of the data matrix columns into libraries, see lib.factor

data <- simulateMPRA(tr = rep(2,10), da=NULL, nbatch=2, nbc=20)
obj <- MpraObject(dnaCounts = data$obs.dna, 
                  rnaCounts = data$obs.rna, 
                  colAnnot = data$annot)
obj <- estimateDepthFactors(obj, lib.factor = "batch", which.lib = "both")
## Upper quantile, using a higher quantile than 0.75:
obj <- estimateDepthFactors(obj, lib.factor = "batch", which.lib = "both",
                            depth.estimator = function(x) quantile(x, .95, 
                                 na.rm=TRUE))