[BioC] DESeq: how to get the variance information

Simon Anders anders at embl.de
Sat Jan 14 09:33:16 CET 2012


On 2012-01-14 07:48, Yuan Tian wrote:
> I'm using DESeq to analyze the RNA-seq data, and I would appreciate
> if anyone can tell me how to get the estimated variation from the
> dataset (in addition to the dispersion value, I would like to check
> the variance as well). It seems that we can use scvPlot to get the
> variance information, but this function has been removed in the
> current DESeq version...

The relation between dispersion and variance is

    variance = mean + dispersion * mean^2

This way, you can recover the variance. Note, however, the DESeq 
performs an additional bias correction that you miss in this simple way.

   Simon



More information about the Bioconductor mailing list