This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How To Export Normalized Counts From Edger

Is normalized counts stored in "$pseudo.alt" after estimating common dispersion?

## edgeR run script
library(edgeR)
setwd("C:/Documents and Settings/user/My Documents/")

# define input Expression file name to read
input <- "test.exp.txt"

# input Expression file
raw.data <- read.delim(input)

# read the sample columns
d <- raw.data[,2:204]  # Edit it accordingly

# read row or gene names
rownames(d) <- raw.data[,1]

# define the groups
group <- c(rep("A",68),rep("B",65),rep("C",70))  # Edit it accordingly

##edgeR stores data in a simple list-based data object called a DGEList. The function readDGE makes a DGEList object directly.
d <- DGEList(counts = d, group=group)

## By default, calcNormFactors uses the TMM method and the sample whose 75%-ile (of library-scale-scaled counts) is closest to the mean of 75%-iles as the reference.
d = calcNormFactors(d)

##To estimate common dispersion:
## The common dispersion is the "squared coefficient of variation", where the coefficient of variation gives the amount of variability in the true abundance for each gene between replicates
#The square root of the common dispersion gives the coefficientcient of variation of biological variation (BCV)#
d = estimateCommonDisp(d, verbose=TRUE)

# Normalized counts ???
d$pseudo.alt
# export Normalized counts table to a file 
write.table(d$pseudo.alt, file="TMM_edgeR_Normalized.test.exp.txt", row.names = TRUE, col.names = TRUE, sep = "\t" )
edger counts

Can't you simply test this explicitly by plotting the data against the input counts? Or examining the summary characteristics (i.e. mean, median)?

I have been having the same question and found the same answer using edgeR with the "usual" method.

But when I started using the GLM method, I discovered that I couldn't find the "pseudo.counts" object anymore after estimating the common dispersion.

Anyone knows where I can find this information in the GLM way ?

Could you post this as a new question?

Yes sure, sorry, I hesitated about the good practice.

No worries, I'll go ahead and delete your answer here then.

I am having the same issue, I don't understand how to extract normalized count table after finding calcNormFactors, as the manual says 'users are advised not to interpret the psuedo-counts as general-purpose normalized counts

How did you resolved this?​'.

3 answers

It has been clearly mentioned in the edgeR guide that:

Although the idea of the pseudo-counts is to equalize the library sizes, pseudo-counts do not have a simple interpretation as normalized counts because they depend on the experimental design as well as on the library sizes. We do not recommend that they are used for other purposes

Regards,
Ganga Jeena

@Ganga Jeena: thanks for the information...

@Rm

Is there any possibility to extract RLE normalized read counts from raw read counts with edgeR? If so then can you tell me which should be the structure to output then?

This should be posted as a new question.

@Rm

What is the normalized counts tag now? How will I get an output of the normalized count matrix now with edgeR? d$pseudo.counts does not work now it seems?

Yes, pseudo.alt is the normalised count. I remember checking for the same. However you should consider estimating tag-wise dispersions as well.

@Arun; thanks...

It's d$pseudo.counts in the latest version of edgeR (3.0.0)

Log in to answer this question.