[BioC] GO terms not found in goTools and GO

[Seth Falcon]

Wolfgang Huber <huber at ebi.ac.uk> writes:

> Dear Lena,
>
> the point to realize is that biomaRt queries an online database (you
> didn't tell us which one you use), while the GO package is assembled and
> packaged by friendly curators in Seattle every couple of months.
>
> According to this site
> http://www.ebi.ac.uk/ego/DisplayGoTerm?id=GO:0060070
> your term was last modified 2007-03-16, which could explain why it is
> not yet in the latest GO package (version 1.17.0).
>
> If you want an consistent set of GO terms throughout, please use the
> humanLLmappings (soon to be humanEGmappings) package for gene-GO
> associations, not biomaRt. If you use biomaRt, you need to spend more
> effort dealing with the dynamic nature of genome annotation (this is not
> our fault, this is the state of the science).

Good advice.  I want to clarify one detail:

We are working on new Entrez Gene based organism specific annotation
packages for human, mouse and rat.  They are intended as New And
Improved versions of <organism>LLMappings.  We have not finalized how
we will name them, and almost surely they will not be humanEGmappings.
I will start a thread on bioc-devel today with some proposed naming
schemas for those that are interested.

Best,

+ seth

-- 
Seth Falcon | Computational Biology | Fred Hutchinson Cancer Research Center
http://bioconductor.org