[BioC] gcrma and other normalizations

[Remo Sanges]

On Jun 2, 2004, at 7:09 AM, Raffaele Calogero wrote:

> Hi,
> if a don't mistake the present implementation of gcrma uses directly  
> the quantile normalization.
> Is there any command to switch off the quantile normalization, or how  
> can I modify the gcrma function to eliminate the quantile  
> normalization?
>
> Since I would like to see  the effect of other normalization  
> procedures on gcrma, is it a risonable procedure to normalize the data  
> first  and subsequently applying the gcrma intensity calculation,  
> after having modified the function in orther to remove the quantile  
> normalization?
> data<-ReadAffy()

Perhaps I am not update (I am not yet passed to the 1.4) but it seems  
to me that
you should use expresso in order to get summary values if you want to  
combine
different type of algorithms.
Furthermore if I don't wrong you should apply background correction
before normalization.

data<-ReadAffy()

data.gcrma<- 
bg.correct.gcrma(data,gcgroup=getGroupInfo(data),estimate="mle")

data.gcrma.invariant<-normalize(data.gcrma, method="invariantset")

summary<- 
expresso(data.gcrma.invariant,bg.correct=FALSE,normalize=FALSE,pmcorrect 
.method="pmonly",summary.method="medianpolish")


Remo