[BioC] Human Gene 1.1 ST cdf

[James W. MacDonald]

Hi Nicolas,

On 3/21/2011 12:51 PM, Nicolas Servant wrote:
> Thanks Jim.
> I found a cdf on BrainArray
> http://www.affymetrix.com/browse/products.jsp?navMode=34000&productId=131554&navAction=jump&aId=productsNav#1_1
> Then I plan to simply use the RMA function from the affy package.
>
> require(hugene11stv1hsentrezgcdf)
> require(hugene11stv1hsentrezg.db)
> require(hugene11stv1hsentrezgprobe)
> cdfname<- "hugene11stv1hsentrezg"
> abatch<- ReadAffy(celfile.path=rawDataDir, cdfname=cdfname)
> rmaData<- rma(abatch, cdfname = cdfname)
>
> Is there still some reason to use xps or oligo instead of the affy
> package ?

It depends on what you are trying to do.

The MBNI re-mapped cdfs are nice because you get updated mappings, but 
for these chips the original mappings aren't that out of date. So I 
don't know how much you gain. In addition, AFAIK, the probes are 
collected into gene or transcript-level probesets, so you cannot do any 
exon-level analyses.

Both oligo and xps will do RMA, but they also allow you to vary at what 
level you collapse the probes into probesets (exon or gene level). If 
you are interested in alternate splicing, then you might want to be able 
to do exon-level stuff. The downside of oligo and xps is that they take 
Affy at their word vis a vis the mapping of probes to (exons, 
transcripts), which involves mapping some probes to multiple things. 
This is a bit sketchy; by definition you don't know what those probes 
are measuring.

The MBNI cdf packages have no multiple-mapped probes (they are 
discarded), so that isn't an issue.

Best,

Jim




> The vignette of oligo seems to also use RMA ...
> http://www.bioconductor.org/packages/2.7/bioc/vignettes/oligo/inst/doc/V5ExonGene.pdf
>
> Thanks
> Nicolas
>
> James W. MacDonald a écrit :
>> Hi Nicolas,
>>
>> On 3/21/2011 11:18 AM, Nicolas Servant wrote:
>>> Hi,
>>>
>>> I'm looking for the cdf package of the Affymetrix Human Gene 1.1 ST
>>> ArrayPlate.
>>> I found the hugene1.0 package but not the 1.1 (only the annotation
>>> package .db are available) ...
>>> Does anybody know if such a package exists ?
>>
>> No. We build the cdf package for this chip (1.0) using the unsupported
>> cdf file that affy supplies. They have never to my knowledge offered
>> an unsupported version of the 1.1 chip, so we don't have a cdf package.
>>
>> You can however use some functions from the aroma.affymetrix project
>> to create a cdf if you are feeling adventurous
>> (https://stat.ethz.ch/pipermail/bioconductor/2010-December/036730.html).
>>
>> But you should really be using either oligo or xps to analyze these
>> chips. The affy package was never intended for these chips, and there
>> are real reasons not to use it.
>>
>> Best,
>>
>> Jim
>>
>>
>>>
>>> Thanks
>>> Nicolas
>>>
>>
>
>

-- 
James W. MacDonald, M.S.
Biostatistician
Douglas Lab
University of Michigan
Department of Human Genetics
5912 Buhl
1241 E. Catherine St.
Ann Arbor MI 48109-5618
734-615-7826
**********************************************************
Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues