Marc, Thank you very much for the information. Does it mean that I have to use those acceptable "evidence code" to GOstats package or I can have my personal label? Thanks. X.G. Date: Tue, 22 Jan 2013 10:23:45 -0800 From: Marc Carlson <mcarlson@fhcrc.org> To: bioconductor@r-project.org Subject: Re: [BioC] question about "evidence code" in data.frame from     GOstats packages with unsupported model organisms Message-ID: <50FED931.60706@fhcrc.org> Content-Type: text/plain Hi, The evidence codes are defined by the folks who work on GO, and it refers to the kind of evidence that supports applying a particular GO term to a specific gene ID.  For many genes to GO mappings in non- model organisms the association is "second hand", meaning that it was really studied on another organism and then assumed to also apply in this other organism (often because of sequence homology). Here is a web site that lists some acceptable evidence codes: http://www.geneontology.org/GO.evidence.shtml Hope this helps,   Marc On 01/21/2013 09:07 AM, Fix Ace wrote: > Just realized some typo in the subject of previous post. So would like to raise this question again since nobody answer it last time. This is relevant to usage of GOstat package with unsupported model organism. > > Thanks. > > > ________________________________ > > To: "bioconductor@r-project.org"<bioconductor@r-project.org> > Sent: Wednesday, January 16, 2013 9:38 AM > Subject: question about "evidence cole" in data.frame from GOstats packages with unsupported model organisms > > > I am planning to use read.table to create the objects that feed to GOFrame. The manual says that the three columns are required, GO IDs, evidence code and gene IDs. I think I understand the first and third column, but what information should I put in evidence code? > > For example I have the following in text file: > --------------------------------- > > GO:0002576    ***    gene1 > > ------------------------------- > > Does it work? >     [[alternative HTML version deleted]] > > > > _______________________________________________ > Bioconductor mailing list > Bioconductor@r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor     [[alternative HTML version deleted]] [[alternative HTML version deleted]]

Hi, I don't recall if the code checks the validity of the evidence codes, but I actually hope that it does. That is, I do think that you *should* be using standard evidence codes. Not because I think GO has thought of everything, but just because if you follow existing standards like this it makes all your work more universal and hence more useful in the future. Also, if you have an important new kind of evidence code that GO is not currently supporting, I would hope that you would contact them and tell them about it. Marc On 01/24/2013 08:22 AM, Fix Ace wrote: > > Marc, > > Thank you very much for the information. Does it mean that I have to use those acceptable "evidence code" to GOstats package or I can have my personal label? > > Thanks. > > X.G. > > > > > > Date: Tue, 22 Jan 2013 10:23:45 -0800 > From: Marc Carlson<mcarlson@fhcrc.org> > To: bioconductor@r-project.org > Subject: Re: [BioC] question about "evidence code" in data.frame from > GOstats packages with unsupported model organisms > Message-ID:<50FED931.60706@fhcrc.org> > Content-Type: text/plain > > Hi, > > The evidence codes are defined by the folks who work on GO, and it > refers to the kind of evidence that supports applying a particular GO > term to a specific gene ID. For many genes to GO mappings in non- model > organisms the association is "second hand", meaning that it was really > studied on another organism and then assumed to also apply in this other > organism (often because of sequence homology). > > Here is a web site that lists some acceptable evidence codes: > > http://www.geneontology.org/GO.evidence.shtml > > > Hope this helps, > > > Marc > > > > > On 01/21/2013 09:07 AM, Fix Ace wrote: >> Just realized some typo in the subject of previous post. So would like to raise this question again since nobody answer it last time. This is relevant to usage of GOstat package with unsupported model organism. >> >> Thanks. >> >> >> ________________________________ >> >> To: "bioconductor@r-project.org"<bioconductor@r-project.org> >> Sent: Wednesday, January 16, 2013 9:38 AM >> Subject: question about "evidence cole" in data.frame from GOstats packages with unsupported model organisms >> >> >> I am planning to use read.table to create the objects that feed to GOFrame. The manual says that the three columns are required, GO IDs, evidence code and gene IDs. I think I understand the first and third column, but what information should I put in evidence code? >> >> For example I have the following in text file: >> --------------------------------- >> >> GO:0002576 *** gene1 >> >> ------------------------------- >> >> Does it work? >> [[alternative HTML version deleted]] >> >> >> >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor@r-project.org >> https://stat.ethz.ch/mailman/listinfo/bioconductor >> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor > > [[alternative HTML version deleted]] > [[alternative HTML version deleted]] > > > > _______________________________________________ > Bioconductor mailing list > Bioconductor@r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor [[alternative HTML version deleted]]