check how normalization of count data worked

Question

[tonja.r]

In DESeq one can normalize the read counts using estimateSizeFactor and then look at the dispersion of the gene using estimateDispersion. The normalized counts can be extracted with counts(cds,normalized=TRUE). And the dispersion graph with the plotDispEsts.

I have other normalization methods (for instance edgeR) and I would like to check for the dispersion there or to check how the normalization of the data worked. What is the best way to show it? I do not think that plotting the distribution of the normalized counts for each sample for each normalization would be comparable somehow.

[h.mon]

If you have your data on an ExpressionSet object, you can use boxplot() on it.