[BioC] which method in the fonction normalize()

Laurent Gautier lgautier at altern.org
Fri Oct 8 22:42:06 CEST 2004


F Duan wrote:
> Default should be "quantile" if I remember correct.

That's right.

To be precise, it is the default in:

getOption("BioC")$affy$normalize.method

(this was done for a future-but-yet-to-come scheme to let one specify 
default methods in a R config file ... possibly the setHook methods can 
let one do so, no time to investigate it... there is also few lines 
about the package options in the affy vignette about cdfenvs).


 > Quoting gregory voisin <voisingreg at yahoo.fr>:
 >>which difference betwen
 >>datanormal<-normalize(data)
 >>datanrormal<-rma(data)

'normalize' just performs the 'normalization' step (or whatever name you 
may want to give it) at the probe level.
'rma' does the full monty: from raw probe level data to probe sets 
expression values. This is mainly a handy wrapper for the function 
'expresso' with the set of pre-processing methods proposed by the rma 
authors.




hth,


Laurent



More information about the Bioconductor mailing list