[BioC] TMM normalization of DESeq input data

Ryan C. Thompson rct at thompsonclan.org
Fri Jun 14 01:58:37 CEST 2013

You would not feed the normalized data to DESeq. Instead, you should 
calculate TMM normalization factors using edgeR::calcNormFactors and 
then convert the norm factors to size factors and assign them into the 
CountDataSet, which should contain the same raw counts as you're using 
for edgeR. I believe that size factors are just the normalization 
factors times the library sizes, but you might want to check that.

On Thu 13 Jun 2013 12:18:47 PM PDT, Narges [guest] wrote:
> Hi,
> I have done differential expression analysis over RNA-seq data using both edgeR and DESeq. There are some recently published studies which suggest TMM as the most efficient normalization method (along with DESeq normalization).edgeR already uses TMM. I wanted to ask do you think it is meaningful if I first normalized my raw data using TMM method and send the normalized data to DESeq and of course remove the DESeq normalization commands and then comparing the results obtained from edgeR.
> Actually the idea is to have same normalization method for both packages and then comparing the DE results.
> Thank you in advance.
>   -- output of sessionInfo():
> R version 2.15.1 (2012-06-22)
> Platform: x86_64-pc-mingw32/x64 (64-bit)
> --
> Sent via the guest posting facility at bioconductor.org.
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

More information about the Bioconductor mailing list