[BioC] RMA() versus expresso with RMA options

[James MacDonald]

I think you would have to use normalize.method=quantiles to get rma. Also, using bg.correct=rma2 gives results similar to what you would get with Bioconductor 1.1. If you want results similar to rma in 1.2, you have to use bg.correct=rma.

You can see the details of what bg.correct=rma does by looking at the functions bg.adjust and bg.parameters. Bg.correct=rma is compiled code, so it is more difficult to inspect.

Jim



James W. MacDonald
UMCCC Microarray Core Facility
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109
734-647-5623

>>> "cmprobst" <cmprobst at terra.com.br> 05/02/03 12:46PM >>>
Hi, everybody,


I have analyzed a set of CEL files with rma() and with expresso(), using the following options: normalize.method=qspline, bg.method=rma, pmcorrect.method=pmonly, summary.method=medianpolish.

As probably some people are aware, the expression intensity values obtained from both methods are not identical.

What is the reason? Maybe RMA() uses rma2 as the method to correct for background values? If so, what is the difference between these two methods of background correction?

TIA

Christian
	[[alternate HTML version deleted]]

_______________________________________________
Bioconductor mailing list
Bioconductor at stat.math.ethz.ch 
https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor