[BioC] Odd isolated problem with GOstats

[Marc Carlson]

Mark W Kimpel wrote:
> I have used a custom wrapper function to invoke GOstats for almost a 
> year without problem. Now, with just one dataset, I am getting an error. 
> Not sure if this is a bug or something very odd about just my one 
> dataset. The GO:ID that yields the error is a valid ID, I checked. I am 
> using the devel version and rechecked after just updating my packages.
>
> Any ideas as to what the problem might be?
>
> Loading required package: rat2302
> Error in .checkKeys(value, Lkeys(x), x at ifnotfound) :
>    invalid key "GO:0016089"
>
> Enter a frame number, or 0 to exit
>
>   1: limma.contrast.output.func(AOP, fit2, rslt, contrast.num = 1, 
> custom.contra
>   2: GO.anal.func(input.df = t.tab.annot.obj$sig.named.genes, 
> annotationPckg = a
>   3: GOHyperMaxCats.func(optimizedParam)
>   4: data.frame(summary(test.obj))
>   5: summary(test.obj)
>   6: summary(test.obj)
>   7: .local(object, ...)
>   8: sapply(mget(goIds, GOenv("TERM")), Term)
>   9: lapply(X, FUN, ...)
> 10: is.vector(X)
> 11: mget(goIds, GOenv("TERM"))
> 12: mget(goIds, GOenv("TERM"))
> 13: `keys<-`(`*tmp*`, value = c("GO:0007268", "GO:0007214", 
> "GO:0050877", "GO:0
> 14: `keys<-`(`*tmp*`, value = c("GO:0007268", "GO:0007214", 
> "GO:0050877", "GO:0
> 15: switch(as.character(direction(x)), "1" = `Lkeys<-`(x, value), "-1" = 
> `Rkeys
> 16: `Lkeys<-`(x, value)
> 17: `Lkeys<-`(x, value)
> 18: .checkKeys(value, Lkeys(x), x at ifnotfound)
>
>
>  > sessionInfo()
> R version 2.7.0 Under development (unstable) (2007-12-20 r43739)
> x86_64-unknown-linux-gnu
>
> locale:
> LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=en_US.UTF-8;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C
>
> attached base packages:
> [1] splines   tools     stats     graphics  grDevices utils     datasets
> [8] methods   base
>
> other attached packages:
>   [1] rat2302_2.0.1        affycoretools_1.11.2 annaffy_1.11.1
>   [4] KEGG_2.0.1           GO_2.0.1             gcrma_2.11.1
>   [7] matchprobes_1.11.0   biomaRt_1.13.6       RCurl_0.8-3
> [10] GOstats_2.5.0        Category_2.5.0       genefilter_1.17.7
> [13] survival_2.34        RBGL_1.15.7          annotate_1.17.3
> [16] xtable_1.5-2         GO.db_2.1.1          AnnotationDbi_1.1.8
> [19] RSQLite_0.6-4        DBI_0.2-4            graph_1.17.14
> [22] limma_2.13.3         affy_1.17.3          preprocessCore_1.1.5
> [25] affyio_1.7.9         Biobase_1.17.8
>
> loaded via a namespace (and not attached):
> [1] cluster_1.11.9 XML_1.93-2
>   
>From looking at the latest package, it appears that this term was just
not in the version of GO that was used to generate the latest GO and
data packages (the data for this was last updated in the Fall). You can
expect an updated version of GO will be available for all packages in
the spring as part of our biannual update process. We update the
packages in devel as we make bug fixes etc. But we freeze the data
sources twice a year (and especially GO because everything depends on
it) so that they are uniform across the datasets provided by ourselves
and our contributors.

I hope this helps,

Marc