[BioC] use of huex10stv2cdf_1.15.0.tar.gz and human exon array analysis

James W. MacDonald jmacdon at med.umich.edu
Tue Jan 27 14:39:17 CET 2009

Hi Dan,

Dan Strumpf wrote:
> Hello,
> I am new to exon array analysis and have used the huex10stv2cdf_1.15.0.tar.gz 
> package (downloaded from 
> http://aurora.lbl.gov/static/huex10stv2cdf_1.15.0.tar.gz
> )as a cdf source/mapping for processing Human exon array data within 'affy' 
> package. Processing went smoothly without any error messages. Noticed some 
> communication about this package in Oct. 2007, and am not clear if results 
> using it are relaible.

I don't know anything about that cdf, but if it is similar to the 
modified cdf that you can get from PICR for exonmap, you should get 
pretty much the same values, as exonmap uses the affy package and a 
modified cdf to compute expression values.

The upside to using exonmap is that you have ways to map the probesets 
to their location on the genome, and could thereby filter out those that 
hit what are considered to be inter-genic or intronic positions.

I don't know if you will be able to use gcrma, as there are no MM probes 
on the exon chips. Instead they just have two sets of probes that are 
designed to measure background.

> I am interested in processing Human exon array data also with gcRMA and am 
> wondering if should use the above package or use xps/exonmap/oligo instead? 
> Would like to know also what are to options for differential expression 
> analysis equivalent to samr limma or RankProd for paired samples from within 
> exon array analysis packages or could the preprocessed/normalized data be 
> exported and used with xps/exonmap/oligo packages.

Pre-processing and differential expression are in general handled by 
different packages. For instance, you don't use the affy or gcrma 
package to look for differentially expressed genes, and the same holds 
true for xps, exonmap, and oligo.

That said, there isn't really that much different between the exon chip 
data and the 3' biased arrays, so tools like limma and RankProd should 
be fine for analyzing these data.



> Thanks,
> Dan
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

James W. MacDonald, M.S.
Hildebrandt Lab
1150 W. Medical Center Drive
Ann Arbor MI 48109-5646

More information about the Bioconductor mailing list