Hi We have re-annotated the ath1 probes with the V6 annotation for
Arabidopsis. You can find it here
in this probe
25mers with multiple gene matches are excluded.. We use probe level
modeling for gene expression estimates. This is likely more than you
but I thought I put it out there as my solution to the problem..
Date: Thu, 17 Aug 2006 15:01:24 +0200
From: Bj?rn Usadel <firstname.lastname@example.org>
To: Nianhua Li <nli at="" fhcrc.org="">
Cc: bioconductor at stat.math.ethz.ch
Message-ID: <44E468A4.7050101 at mpimp-golm.mpg.de>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
as Tine pointed out I would suggest you choose multiple....
or you do it the hard way. For our purposes (MapMan visualization
on classification) I test, if the multiple genes hit by one probeset,
have a similar function. If this is the case I mix the annotations
assuming that it might be a [diverged] gene family, in which case
might be some information left (Affy used to tag them _s_, but affy is
way outdated) when I sample the whole class.
However, if a probesets turns out to hit genes of different classes
gene families ancient _x_ tag] (e.g. glycolysis and say proteasom
dependent degradation) I annotate the probeset as "hitting multi" and
put it in a special "non-evaluate able" class.
You could also try to determine if it is really a gene family that is
hit, in which case the annotations would be similar as well anyway.
But that is a lot of querying and eventually needs manual interaction.
Thanks for your work.
Nianhua Li wrote:
> Hi, Tine, Bjorn, Thomas and other Arabidopsis experts,
> Thanks a lot for the feedbacks. I will get the update done this week
> could help me to solve the following problem :P
> In TAIR's probe-to-locus mapping file, for example
> some probesets are mapped to >= 1 locus. However, in annotation
> ath1121501 and ag, all annotations (e.g. agCHRLOC, agENZYME) are
> probeset identifier. It assumes a one-to-one mapping between
> so that the annotation to a gene is the annotation to a probeset.
> How to handle the one probeset to multiple locus mappings? I can
> possible solutions:
> 1. pick the "best" locus, but how?
> 2. mix the annotations to all mapped locus together
> 3. set to NA
> Any suggestions are highly appreciated. Many thanks!
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> Search the archives:
Bj?rn Usadel, PhD
Max Planck Institute of Molecular Plant Physiology
System Regulation Group
Am M?hlenberg 1
Tel (+49 331) 567-8114
Email usadel at mpimp-golm.mpg.de