[BioC] problems with annotation package (error in AnnotationDbi v1.18.1 ?)

[Marc Carlson]

Hi Sam,

There are a couple of things going on here.  The 1st is that you are 
using the same package AGILENTv2.db with both bioc 2.10 AND also with 
bioc 2.9.  That is guaranteed to be a bad idea, because each version of 
bioconductor (including the annotation packages is tested with its set 
of same version software packages.

Now the AGILENTv2.db package does not look like one that you got from 
us, so I can't tell you which version of bioc you should be using it 
with, that would depend on which version it was made by.  But I can tell 
you that it should probably not be used with any other versions (other 
than the one it was made with).

The second thing that is going on is that a chipDb package like 
AGILENTv2.db appears to be does not really have much annotation data in 
it.  If you look inside of the DB that is wrapped in a chipDb package, 
you will notice that there is not really much data in it.  That is 
because packages like this point to an organism package like 
org.At.tair.db for most of their data.  The organism package does have a 
lot of data in it, and is joined to the chipDb package via gene 
identifiers to get the data that you are looking at in your example.  So 
the fact that you have two different versions of bioc installed that 
each have their own version matched organism packages is causing the 
differences that you are seeing.

So if you really wanted to make the results be the same, you would need 
to use the same version of the org packages in both cases.  But please 
don't do that!  Instead, please remember that those packages are made 
for each version of bioc.  Because of the testing limitations inherent 
to a system where all the packages may change for every single release, 
we can only test to make sure that they work with the other packages 
within that specific version of bioc.  If you mix and match packages 
across different versions, it is possible, and even somewhat likely that 
you will get some unexpected results.

Please let me know if you have more questions,


   Marc




On 06/14/2012 10:55 AM, Samuel Wuest wrote:
> Hi all,
>
> I have a problem that concerns the use of annotation packages, here in an
> example for a custom-made AGILENT microarray for Arabidopsis (but I think
> it also concerns the org.At.tair.db package). I don't quite understand how
> this issue arises, so sorry for the vague title (my guess is that it
> concerns AnnotationDbi package version 1.18.1).
>
> The issue is: if I query a GO-ID from my AGILENTv2GO2ALLPROBES bimap, I get
> a very different result using a newer BioC version when compared to the
> result using an older BioC version. Please note here that the annotation
> package itself is the same in both versions used!
>
> Here is the output from a query in the NEW version:
>
>
>> library(AGILENTv2.db)
>> genes<- get("GO:0006351", AGILENTv2GO2ALLPROBES)
>> length(unique(genes))
> [1] 1825
>
> #### ok, so that would be 1825 unique genes/probes obtained from the query
>
>> sessionInfo()
> R version 2.15.0 (2012-03-30)
>
> Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
>
>
> locale:
>
> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
>
>
> attached base packages:
>
> [1] stats     graphics  grDevices utils     datasets  methods   base
>
>
> other attached packages:
>
> [1] AGILENTv2.db_2.6.4   org.At.tair.db_2.7.1 RSQLite_0.11.1
> DBI_0.2-5            AnnotationDbi_1.18.1
>
> [6] Biobase_2.16.0       BiocGenerics_0.2.0
>
>
> loaded via a namespace (and not attached):
>
> [1] IRanges_1.14.3 stats4_2.15.0
>
> ------------------------------------------
> Now please compare this with the output from the OLD version.
>
>> library(AGILENTv2.db)
>> genes<- get("GO:0006351", AGILENTv2GO2ALLPROBES)
>> length(unique(genes))
> [1] 2122
> #### here, there are 2122 unique genes/probes obtained, so many more than
> the 1825 above, even though the AGILENTv2.db used was the same (version
> 2.6.4)
>> sessionInfo()
> R version 2.14.0 (2011-10-31)
> Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
>
> locale:
> [1] en_IE.UTF-8/en_IE.UTF-8/en_IE.UTF-8/C/en_IE.UTF-8/en_IE.UTF-8
>
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods   base
>
> other attached packages:
> [1] AGILENTv2.db_2.6.4    org.At.tair.db_2.6.4  RSQLite_0.11.1
> DBI_0.2-5             AnnotationDbi_1.16.10
> [6] Biobase_2.14.0
>
> loaded via a namespace (and not attached):
> [1] IRanges_1.12.5
>
> -------------------------
>
> Any suggestions? I am not sure in which package the problem is, because the
> annotation package should be identical. I think the same problem occurs in
> the org.At.tair.db package, however, I have also different versions of the
> annotation package installed on the different computers...
>
> Thanks for any help,
>
> Sam
>