This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Logconc In Edger

Hi,

I am running a RNA-Seq analysis with edgeR. I have the results in three columns. I get the meaning of fold-change and p-value. can someone please explain to me in "simple" words, what is the meaning of the logConc? and maybe how it is being calculated??

Thanks Assa

edger rna

2 answers

From the edgeR user manual: logConc, the log-average concentration/abundance for each tag in the two groups being compared.

This is the log of the average abundance of the sequence read between control and experimental.

Thanks Larry, I read the manual, but I don't exactly understand, what that means. Is there a way of getting an example with numbers? I'll tell you what I have as a result. For one of my experiments I have logConc of ~-34.5, with a foldFC of 31. The read counts are for the control 71 and for the mutant 0. Can you explain to me how this is being calculated?

No, I cannot, especially with a mutant read count of 0. log(x/0) is undefined.

To view the calculation of the function, run the functionname without arguments and parentheses in R. For the exactTest, the relevant calculation steps for logConc are:

pair        <- levels( as.factor( object$samples$group ) )[1:2]
this.pair   <- (object$samples$group %in% pair)
lib.size    <- object$samples$lib.size * object$samples$norm.factors
group.pair  <- factor( as.vector( object$samples$group[this.pair] ) )
obj.pair    <- DGEList(counts = object$counts[, this.pair], group = group.pair, lib.size = lib.size[this.pair])
q2q.pair    <- equalizeLibSizes(obj.pair, disp = dispersion, null.hypothesis = TRUE)
levs.pair   <- levels(group.pair)

logConc <- (log2(q2q.pair$conc$conc.group[, pair[1] == levs.pair]) + 
            log2(q2q.pair$conc$conc.group[, pair[2] == levs.pair]))/2

The object is of the class DGEList. Hope this helps.

Log in to answer this question.