fpm: FPM: fragments per million mapped fragments

In aghozlane/DESeq2shaman: Differential gene expression analysis based on the negative binomial distribution

Description

Calculates either a robust version (default) or the traditional matrix of fragments/counts per million mapped fragments (FPM/CPM). Note: this function is written very simply and can be easily altered to produce other behavior by examining the source code.

Usage

fpm(object, robust = TRUE)

Arguments

object	a DESeqDataSet
robust	whether to use size factors to normalize rather than taking the column sums of the raw counts.

Value

a matrix which is normalized per million of mapped fragments, either using the robust median ratio method (robust=TRUE, default) or using raw counts (robust=FALSE).

See Also

fpkm

Examples

# generate a dataset with size factors: .5, 1, 1, 2
dds <- makeExampleDESeqDataSet(m = 4, n = 1e3,
                               interceptMean=log2(1e3),
                               interceptSD=0,
                               sizeFactors=c(.5,1,1,2),
                               dispMeanRel=function(x) .01)

# examine the column sums
# then add 1 million counts over rows 11:15 for each sample
colSums(counts(dds))/1e6
counts(dds)[11:15,] <- 2e5L
colSums(counts(dds))/1e6

# the robust FPM treats the samples
# relatively equally
head(fpm(dds), 3)

# the non-robust version is thrown
# off by the 5 rows with large counts
head(fpm(dds, robust=FALSE), 3)

# the column sums of the robust version
# are not equal to 1e6, but the
# column sums of the non-robust version
# are equal to 1e6 by definition
colSums(fpm(dds))/1e6
colSums(fpm(dds, robust=FALSE))/1e6

# the total sum is equal for both methods
sum(fpm(dds))
sum(fpm(dds, robust=FALSE))

aghozlane/DESeq2shaman documentation built on Nov. 1, 2019, 9:01 p.m.