DESeq: question about baseMean

[Azazel]

In DESeq, what exactly is baseMean and how is it computed?

My guess is that baseMean is a kind of normalized tag count (however if that's correct they would probably call it something like tagMean or exprMean, not baseMean?!)

I read "Analysing RNA-Seq with the DESeq package", version from April 21, 2011 but can not find an explanation of baseMean.

[Simon Anders]

Yes, it is the normalized count, i.e., simply the number of reads divided by the size factor (normalization constant) of the sample. I know that the name is not well chosen; this is why the term does not appear in the paper any more.

[Azazel]

Hi Simon, thank you for your answers in this and the other thread. I would like to to suggest to leave the term in the documentation actually, but to explain it. Just from reading the DESeq vignette as it's now I think it's not clear what it is exactly.