[BioC] how to unlock a package environment

Marc Carlson mcarlson at fhcrc.org
Fri Oct 2 01:07:53 CEST 2009


Hi Ariel,

The reason why the arabidopsis chip packages do not have an ENTREZID
mapping is because this is not what these packages are primarily based
upon and we didn't want people to grab those IDs out of habit from using
the other chip packages.  Instead, these packages are based on the
"TAIR" sources which are popular with the community that likes to study
Arabidopsis.  As a result of this, the mapping that you want to use to
filter the results should be the TAIR IDs which for historical reasons
live in the ACCNUM mapping for the arabidopsis packages.  Herve beat me
to the punch, but the nsFilter method in the latest version of the devel
branch has already been modified to determine what kind of annotation
package you are using and then make the correct decision about what kind
of mapping it should be using to filter. 

If you really want entrez gene IDs for something else, you can get them
by using the ENTREZID mapping found in the org.At.tair.db package.  All
of the TAIR IDs from your chip package should have a match in that
package which is also primarily based on the TAIR data sources.

I have made a change to the release version of this package to correct
this bug, it should show up online within about a day.

As for unlocking the packages, we strongly discourage that as it can
cause all manner of odd behavior with other packages that depend on
these annotations to produce standard annotations...  But if you really
want to play with "bombs and guns", you should be able to set the
database file to be write-able.


  Marc



Ariel Chernomoretz wrote:
> Hi all,
> I am currently analyzing some arabidopsis thaliana experiments (ath1121501
> GeneChip)
> The corresponding annotation package, ath1121501.db, is a little bit
> 'special' as
> there is no ENTREZID map for the chip.
> I would like to cheat and create  a fake ath1121501ENTREZID for the
> environment as a copied instance of ath1121501ACCNUM (in order to use, for
> instance,  the genefilter package's function "nsFilter" to deal with non
> annotated, or duplicated probesets)
>
> The problem is that I am not able to create a new entry in the
> package:ath112150.db as it is a locked environment. Is there any way to
> temporarily unlock it, do some stuff and lock it again?
> (It seems easy to lock environments, using the lockEnvironment function, but
> I did not find an easy way to unlock them)
>
> Thanks in advance
> Ariel./
>
>



More information about the Bioconductor mailing list