Jose You weren't being cryptic, the packages are! I have put some code on the use of topGO on our website: http://bioinformatics.iah.ac.uk/sample-code I hope this of use to you and others Thanks Mick ________________________________________ From: J.delasHeras@ed.ac.uk [J.delasHeras@ed.ac.uk] Sent: 08 February 2010 19:29 To: michael watson (IAH-C) Cc: bioconductor Subject: RE: [BioC] GSEA, topGO, GOstats...? what's a good way to look at GO over-representation? Hi Mick, I didn't think I was cryptic at all, but I'm sorry I didn't make it clear enough. I was just looking for a good way to do GO over-representation analysis starting from say entrez gene IDs, without being tied to a particular array definition (if GO ids are needed I can fish them out). I think that my difficulty with some of these packages (topGO included) was generating an annotation package, generating the correct data structure, and the examples being Affy-centric. I'm sure it seems very simple, once I figured it out, but not right now. I have looked a tools like FatiGO and FatiScan etc... but I was never very happy with them. Admittedly this was a while ago and matters may have improved. Sometimes I felt that some genes didn't have a matching GO term despite my knowing that there was GO information for it, and it often took very long for me to get the results. Maybe I'll have a look again, but I'd much rather keep the work in R *if reasonable*: scripts one can reuse, every step is documented, etc etc. Thanks for your comments, I'll look more closely at topGO. Jose Quoting "michael watson (IAH-C)" <michael.watson at="" bbsrc.ac.uk="">: > These all are a little cryptic! > > I have some sample code for topGO that doesn't use AFFY ids, it uses > a dataset that I can't give out, but at least it's not affy. > > I ended up writing my own code to do this that works from > data.frames etc and sucks the latest annotation directly from the > web, rather than using the bioc annotation packages. Some of this > was wrapped into our package CORNA > (http://bioinformatics.iah.ac.uk/software/corna) > > Also, are you devoted to R? If not, then why not use something like > FatiGo? http://www.fatigo.org/ > > Mick > > -----Original Message----- > From: bioconductor-bounces at stat.math.ethz.ch > [mailto:bioconductor-bounces at stat.math.ethz.ch] On Behalf Of > J.delasHeras at ed.ac.uk > Sent: 08 February 2010 16:30 > To: bioconductor > Subject: [BioC] GSEA, topGO, GOstats...? what's a good way to look > at GO over-representation? > > > Dear list, > > I have a few gene lists derived from a human Illumina expression > array. I just have Illumina IDs, I have gene names, and I have entrez > gene IDs I obtained for them. > > I would like to analyse the list to look for over-representation of > some category, probably using gene ontologies. > I see there are several packages that seem to address this, although > when I look at the examples I get the feeling they were designed with > Affy arrays in mind and depend on an Affy array design... > > I am sure I am not the only one wanting to do this type of work on > non-Affy arrays... I would appreciate a nudge towards the right > package, or a way to "persuade" it to work with non-Affy array data, > after all I imagine that all the array design is used for is the > definition of teh genelists/universe and retrieval of the relevant GO > ids. > > Thank you for any helpful comments. > > Jose > > -- > Dr. Jose I. de las Heras Email: J.delasHeras at ed.ac.uk > The Wellcome Trust Centre for Cell Biology Phone: +44 (0)131 6513374 > Institute for Cell & Molecular Biology Fax: +44 (0)131 6507360 > Swann Building, Mayfield Road > University of Edinburgh > Edinburgh EH9 3JR > UK > ********************************************* > NEW EMAIL from July'09: nach.mcnach at gmail.com > ********************************************* > > -- > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > _______________________________________________ > 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 > > -- Dr. Jose I. de las Heras Email: J.delasHeras at ed.ac.uk The Wellcome Trust Centre for Cell Biology Phone: +44 (0)131 6513374 Institute for Cell & Molecular Biology Fax: +44 (0)131 6507360 Swann Building, Mayfield Road University of Edinburgh Edinburgh EH9 3JR UK ********************************************* NEW EMAIL from July'09: nach.mcnach at gmail.com ********************************************* -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

Hi Jose, If you would rather use a more universal ID such as an Entrez Gene ID, you can use GOstats, and just pass in one of our "org" packages. So for example, if you wanted to do it with human and use Entrez Gene IDs, you would do something like this: library("GOstats") params = new("GOHyperGParams", geneIds=yourList, universeGeneIds = yourUniverse, annotation="org.Hs.eg.db", pvalueCutoff=0.05, testDirection="over", ontology = "BP", conditional=FALSE) hOver = hyperGTest(params) report = summary(hOver) Marc J.delasHeras at ed.ac.uk wrote: > > Dear Wolfgang, > > thanks for your reply. Yes, the annotation support part in some of the > vignettes I looked at was throwing me out a bit. > > Jose (no, still haven't changed my name ;-) > > Quoting Wolfgang Huber <whuber at="" embl.de="">: > >> Dear Javier >> >> Chapter 13 of the 'Bioconductor Case Studies' book is a good start, as >> is >> http://www.bioconductor.org/workshops/2009/SeattleApr09/gsea/GSEA_L ecture.pdf >> >> and the vignette of the GSEABase package. >> >> Let yourself not be confused by the fact that in some functions (eg >> GOstats), there is support to make it easier to work with the >> Bioconductor annotation packages (which are provided, among others, for >> Affymetrix genechips). The concept of gene set enrichment analysis >> itself is independent of where you get the gene sets from, and the >> software above works with general gene lists. >> >> And if you do not care so much about automation, reproducibility and >> flexibility of your workflow, then using websites like mentioned by >> Michael to copy-paste your gene lists into might be the way to go. >> >> Best wishes >> Wolfgang >> >> >> J.delasHeras at ed.ac.uk scripsit 02/08/2010 05:29 PM: >>> >>> Dear list, >>> >>> I have a few gene lists derived from a human Illumina expression >>> array. I just have Illumina IDs, I have gene names, and I have >>> entrez gene IDs I obtained for them. >>> >>> I would like to analyse the list to look for over-representation of >>> some category, probably using gene ontologies. >>> I see there are several packages that seem to address this, >>> although when I look at the examples I get the feeling they were >>> designed with Affy arrays in mind and depend on an Affy array >>> design... >>> >>> I am sure I am not the only one wanting to do this type of work on >>> non-Affy arrays... I would appreciate a nudge towards the right >>> package, or a way to "persuade" it to work with non-Affy array >>> data, after all I imagine that all the array design is used for is >>> the definition of teh genelists/universe and retrieval of the >>> relevant GO ids. >>> >>> Thank you for any helpful comments. >>> >>> Jose >>> >> >> >> -- >> >> Best wishes >> Wolfgang >> >> >> -- >> Wolfgang Huber >> EMBL >> http://www.embl.de/research/units/genome_biology/huber/contact > > >