[BioC] DESeq mean,var

[Wolfgang Huber]

Jan/11/12 1:47 PM, Bruce Moran(External) scripsit::
> I am using DESeq for RNAseq data. I am interested in comparing the
> distribution of my samples individually with that of the model used by
> DESeq for all of my data, hoping to rationalise exclusion of several
> samples by this method. Where can I find the estimated mean, variance
> for the model fitted to my data within the output? I assume it is
> created after the nbinomtest function but have dissected the output
> without success. I hope someone can point me in the right direction.
>

Dear Bruce

thanks for the question. In DESeq, mean and variance are computed for 
each gene, but are computed across samples, and are not a per-sample 
property.

If you would like to do per-sample quality control, my suggestion is to 
use variance stabilising transformation (see Section 6 of the vignette) 
and then the arrayQualityMetrics package.

Best wishes
	Wolfgang


Wolfgang Huber
EMBL
http://www.embl.de/research/units/genome_biology/huber