[BioC] 3' 5' ratio

James MacDonald jmacdon at med.umich.edu
Tue Oct 28 16:49:08 MET 2003

It is quite easy to calculate these ratios, but you would not do so
directly from the .cel file (MAS5 doesn't use raw data either). Instead,
calculate expression values using rma, expresso, whatever. Then you can
extract the corresponding values to calculate the ratios.

The probes that are spiked in as molecular biology controls are BioB,
BioC, BioD and CreX. There are two sets of probes for each of the
controls, so you may have to do some digging to see if one or both are
used for the ratio. However, a naive approach would be to simply
calculate the ratios for each chip using each probe, doing something
like this:

Firstchip.BioCs <- exprs(eset)["AFFX-BioB-3_st",1] -
Firstchip.BioCa <- exprs(eset)["AFFX-BioB-3_at",1] -
Secondchip.BioCs <- exprs(eset)["AFFX-BioB-3_st",2] -

etc, etc. 

This will give you the log ratios. You might want to take the antilog
if that is what MAS5 usually reports. You can get all the names of the
control probes by geneNames(eset). They are all near the bottom of the
list.  Here I am using eset as the name of the exprSet you get after
calculating expression values.



James W. MacDonald
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109

>>> Sylvia Merk <merk at ibe.med.uni-muenchen.de> 10/28/03 09:09AM >>>
Hi all,

I have cel-files from Affymetrix U133a+b chips and I wanted to do some

quality control - so I read in the cel-files with the affy-package  and

got an affybatch-object.
Then I had a look at the RNA degradation plots. I know that Microarray

suite performs 3'/5'-ratios as a quality check. My problem is: I've
got the cel-files - there is no other information from Microarray suite

available. So my question is: is it possible to calculate  3'/5'-ratios

out of the cel-file?
Thank you in advance for your help.

Many greetings

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

More information about the Bioconductor mailing list