[BioC] Normalization of array leads to decrease in number of differentially expressed genes

[Kevin R. Coombes]

Normalizing separately is likely to lock in a bias that would be removed 
if you normalize together.  So, I would be more inclined to believe the 
(fewer) differentially expressed genes that you find when normalizing 
together.

On 7/25/2012 6:07 AM, Cytochalasin wrote:
> Hi all,
>
> I performed normalization on my array data (I use qPCR array of 384
> wells), and if I normalize within replicates of all conditions(each
> condition normalized separately) , I get more differentially expressed
> genes that if I normalize all data together., How may I found that
> which result is more accurate?
>
> Thank you in advance
>
> _______________________________________________
> 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