Dear Jim, Marc, list, affymetrix on enquiry: probeset annotations are updated regularly, but the probe to probeset mapping remains unchanged. This means, that they are still based on Drosophila genome build 3.1 and so is bioconductor. I stumbled over customCDFs quite recently, too. Hearing your recommendations and reading through their documentation, they look almost necessary to use. Unfortunately RNAdegradation plots do seem to make difficulties. The problem might be related to the order of probes in the probesets, and which is not entirely clear to me how to circumvent. Anyways thank you for your help so far! David Molnar 2010/1/13 James W. MacDonald <jmacdon at="" med.umich.edu="">: > Hi David, > > I guess it depends on what you are really after. The CDF simply maps the > locations of the probes on a chip back to their probesets, so there is never > any need to update that file (if Affy changed the design of the chip they > would just release a new Drosophila chip as they did when they went to the > Drosophila 2.0 chip). > > I sort of assumed you originally meant to ask what genome build the chip was > designed against, which is what they show in the pdf I directed you to. > Again, this information is set in stone. > > What may change is what the probes are assumed to bind to. If for instance > an EST is later mapped to a particular gene, then the annotation for the > probeset directed against the EST may change to reflect the newer > annotation. These data are contained in the drosophila2.db package, which in > turn is based on the most recent annotation file from Affymetrix. IIRC, Affy > updates these annotation files quarterly, so we would have used the newest > Affy annotation file as of October 2009 for the release version of this > package. > > Note that annotating the probesets is a non-trivial task, and we are simply > taking the probeset -> Entrez Gene ID mappings from Affy and doing the > mapping from Entrez Gene to the various other available annotation data. > Since the annotation of any genome is a moving target, and it is a long time > since 2003, some of the things that the probesets on this chip are intended > to interrogate may no longer be interrogated as well (or at all), according > to the genome as we now know it. Our annotation packages only take this into > account to the extent that the annotation file from Affy does. > > A different approach to this problem is taken by the folks at MBNI, who > create their own CDF files based on the current genome build. If interested, > you can see information about these packages at > > http://brainarray.mbni.med.umich.edu/Brainarray/Database/CustomCDF/g enomic_curated_CDF.asp > > They have set up a CRAN-style repository that biocLite() knows about, so if > interested, you can get their packages directly using biocLite(). > > Best, > > Jim > > > > D?vid Moln?r wrote: >> >> Hi Jim, >> >> thank you for the swift reply. I puzzled over that affymetrix >> datasheet before, and I should probably have been more specific on >> that: has affymetrix never updated this CDF based on Drosophila genome >> release 3.1? Is bioconductor using the unupdated affymetrix CDF? >> Release 3.1 of the Drosophila genome is from 2003. ?Data and gene >> models have considerably improved since then. In the CDF packages on >> the bioconductor server the date of the files is from 2008. I >> speculated that somebody might have had a look on them recently. Is >> this true? >> >> Best regards, >> David >> >> >> 2010/1/12 James W. MacDonald <jmacdon at="" med.umich.edu="">: >>> >>> Hi David, >>> >>> Look at the left margin here: >>> >>> >>> http://www.affymetrix.com/support/technical/datasheets/drosophila2 _datasheet.pdf >>> >>> Best, >>> >>> Jim >>> >>> >>> >>> D?vid Moln?r wrote: >>>> >>>> Dear all, >>>> >>>> I tried to find out in vain from which genome release the current Chip >>>> Description File for GeneChip Drosophila-2 was made. >>>> Could anybody give me please a hint? >>>> >>>> Best regards, >>>> David Molnar >>>> >>>> _______________________________________________ >>>> 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. >>> 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 >>> > > -- > 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 >