calculateCPM: Calculate CPM

In LTLA/csaw: ChIP-Seq Analysis with Windows

Calculate CPM

Description

Calculate counts-per-million (CPM) values for each feature.

Usage

calculateCPM(object, use.norm.factors=TRUE, use.offsets=FALSE, 
    log=TRUE, prior.count=1, assay.id="counts")

Arguments

object	A SummarizedExperiment object containing a count matrix, as produced by windowCounts.
use.norm.factors	A logical scalar indicating whether to use normalization factors, if available.
use.offsets	A logical scalar indicating whether to use offsets, if available.
log	A logical scalar indicating whether log2-transformed CPM values should be returned.
prior.count	A numeric scalar specifying the prior count to add when log=TRUE.
assay.id	A string or integer scalar indicating which assay of y contains the counts.

Details

CPMs are calculated in the standard manner when log=FALSE, use.offsets=FALSE and use.norm.factors=FALSE.

When log=TRUE, a library size-adjusted prior count is added to both the counts and the library sizes, see cpm for details.

When use.norm.factors=TRUE, the effective library size is used for computing CPMs, provided that normalization factors are available in object. This is defined as the product of the library size in object$totals and the normalization factor in object$norm.factors.

If use.offsets=TRUE, the offsets are converted into effective library sizes using scaleOffset. If log=TRUE, this is done after addition of a prior count to both the counts and library sizes, see addPriorCount for details.

Value

A matrix of the same dimensions as object, containing (log-)transformed CPM values for each feature in each sample.

Author(s)

Aaron Lun

See Also

cpm, scaleOffset, addPriorCount

Examples

bamFiles <- system.file("exdata", c("rep1.bam", "rep2.bam"), package="csaw")
data1 <- windowCounts(bamFiles, width=50, filter=1)
head(calculateCPM(data1))

data1$norm.factors <- c(0.5, 1.5)
head(calculateCPM(data1))

head(calculateCPM(data1, log=FALSE))

# With offsets:
assay(data1, "offset") <- matrix(rnorm(nrow(data1)*ncol(data1)), 
    nrow=nrow(data1), ncol=ncol(data1))
head(calculateCPM(data1, use.offsets=TRUE))

head(calculateCPM(data1, use.offsets=FALSE))

LTLA/csaw documentation built on Dec. 21, 2024, 1:10 a.m.