[BioC] GO-Analysis using TopGO for Yeast2 data

[Tobias Koschubs]

Hi!

I tried to do a GO-Analysis using the TopGO package with Affymetrix 
Yeast2 Expression data but encountered a couple of problems. As I'm new 
to Bioconductor, I would be very grateful to get some help.

The Yeast2-Package related problems are the following: Very recently, 
the GO-Annotations have been updated and now include also the 
annotations from computationally predicted methods (available from 
yeastgenome.org and geneontology.org). Of course, I would like to use 
the newest annotations and I tried to use the AnnBuilder 
yeastPkgBuilder-method. However, I didn't succeed to bring the 
gene_association.sgd file together with the Probe mappings on the chip. 
I guess one would need to use the ORFs as identifiers instead of the 
SGD-identifiers from the gene_association file. Moreover, I didn't 
figure out how to set the path for the annotation file as local (it was 
suggested to unpack it in advance). Is there someone from the Biocore 
Data Team who has the proper scripts to do it?

With the TopGO package, I did the GO Analysis using the current yeast2 
annotations. Unfortunately, it's documentation lacks some details and I 
didn't figure out, how to print out the genes (or at least the probes) 
according to the top-GO-terms listed by the genTable-Method. From the 
source-code, I learned that there's a printGenes-method which possible 
should do this. However, the yeast2 chip is a bit different from others 
as it doesn't use ENTREZIDs or SYMBOLS and I also didn't figure out what 
are the parameters of this method.

Best and thanks in advance,
Tobias Koschubs