[BioC] error with GOHyperG

Francois Pepin fpepin at cs.mcgill.ca
Fri Aug 18 17:57:14 CEST 2006


Hi Tony,

I think this is caused by a version mismatch on your annotation
packages. Generally, you want all of them to be made at the same time so
that they're based on the same sources.

Your hgu95av2 annotations are version 1.12.0 while GO is at version
1.8.2. It is very likely that hgu95av2 says a probeset is annotated by a
GO term which your GO annotation knows nothing about, hence the error.

They're separated because the chip annotations only knows about the GO
identifier, while the GO annotation can tell you about its name,
description and position in the hierarchy. It wouldn't make sense to
cause all chip annotations to remember identical information about GO
terms.

I would suggest updating your annotation packages to the same version
and trying again.

Francois

On Thu, 2006-08-17 at 11:28 -0700, Tony Long wrote: 
> All:
> 
> I am getting an error message with GoHyperG (and a similar error with 
> HyperGTable).
> 
> library("hgu95av2")
> w1 <- as.list(hgu95avLOCUSID)
> w2 <- unique(unlist(w1))
> set.seed(123)
> myLL1 <- sample(w2,100)
> myLL2 <- c(myLL1, sample(w2,100))
> xx <- GOHyperG(myLL1,"hgu95av2",what="BP",universe=myLL2)
> 
> Error:  value for 'GO:0031997' not found
> 
> >  sessionInfo()
> Version 2.3.0 (2006-04-24)
> i686-redhat-linux-gnu
> 
> attached base packages:
> [1] "splines"   "tools"     "methods"   "stats"     "graphics"  "grDevices"
> [7] "utils"     "datasets"  "base"
> 
> other attached packages:
> drosophila2cdf    drosophila2        GOstats       Category       hgu95av2
>         "1.4.7"       "1.12.0"        "1.6.0"        "1.4.1"       "1.12.0"
>            KEGG       multtest     genefilter       survival         xtable
>         "1.8.1"       "1.10.2"       "1.10.1"         "2.24"        "1.3-2"
>            RBGL       annotate        Biobase             GO          graph
>         "1.8.1"       "1.10.0"       "1.10.1"        "1.8.2"       "1.10.6"
>           Ruuid
>        "1.10.0"
> 
> Thanks!!
> 
> Tony



More information about the Bioconductor mailing list