[BioC] Expression on different Affy arrays

Park, Richard Richard.Park at joslin.harvard.edu
Thu Jun 17 18:28:06 CEST 2004

Hi Donglei, 
I think using MAS 5 is a perfectly reasonable way to go w/ out being able to use rma to a complete background normalization. I would just be careful about trying to deduce signifigance from genes w/ lower expression and lower fold change values. This kind of analysis would need to be broad. 

The only possible way of combining the 2 different chips into one rma normalization to my knowledge. Would be trying to subset the 2.0 arrays into a format for the 1.0 version. This would require understanding the .cel format and then downloading the individual probeset information (pm and mm values for each x by y pair). It's not an easy task, but it is possible. And also you may not know what can increase the noise into the chips (different number of amplifications, different starting amounts of rna, chip defects, etc.. etc..) 



-----Original Message-----
From: donghu at itsa.ucsf.edu [mailto:donghu at itsa.ucsf.edu]
Sent: Wednesday, June 16, 2004 5:53 PM
To: bioconductor at stat.math.ethz.ch
Subject: [BioC] Expression on different Affy arrays


I have a few MOE430A and MOE430 2.0 arrays.  As you know, MOE430A has
~22000 probe sets and MOE430 2.0 has ~46000 probe sets.  All MOE430A
probe sets are on MOE430 2.0.  Now I need to compare expression
measurement of same probe sets on these 2 types of array.  Is it possible 
to compute RMA measurement of these 2 array types?  If there is no
convenient way to do so, does it sound reasonable to use MAS 5.0 signal
intensity for such comparison as MAS 5.0 uses a target intensity for
normalization?  Thanks for any suggestions.

Donglei Hu

Bioconductor mailing list
Bioconductor at stat.math.ethz.ch

More information about the Bioconductor mailing list