[BioC] retrieveGenomicFeatureAnnotation.R file supplied with Ringo

[Joern Toedling]

Hello,

by default, the biomaRt functions listed in the script query the current
version of Ensembl, and the genome assemblies thus are those used in the
current Ensembl release.

However, it is possible to connect to an archived version of Ensembl which may
contain a previous version of your respective genome build.
For a table of available archived marts, you can use

> listMarts(archive=TRUE)

and then connect to one of those by specifying for example

> ensembl <- useMart("ensembl_mart_43", dataset="hsapiens_gene_ensembl",
archive=TRUE)

and then you should be able to use the rest of the script as with the current
one. There might be minor problems because sometimes attributes get renamed
between Ensembl releases, but usually one can resolve these by looking at

> listAttributes(ensembl)

For checking assembly versions in Ensembl releases, however, I think that you
first need to consult the Ensembl archive pages before picking one. I am not
aware of a more automated way to select the mart based on the desired genome
assembly.

Regards,
Joern

On Thu, 4 Mar 2010 12:10:25 +0000, Paul Geeleher wrote
> Hi,
> 
> I've been following the instructions in the Ringo manual and using
> this file to associate my Chip Enriched regions with genomic 
> features. Thing is it seems that in 
> retrieveGenomicFeatureAnnotation.R biomaRt is querying the HG19 
> genome build. I haven't been able to find a way of changing this to 
> HG18, having searched the web I don't think it's possible? If there 
> is no way of explicitly specifying the genome build maybe it would 
> be worth noting somewhere in the file or documentation that biomaRt 
> is using HG19? Just that I guess the fast majority of people will be 
> using microarrays based on HG18?
> 
> As an alternative I think I'm just going to try reading gene 
> locations from a BED file on the UCSC site.
> 
> Sincerely,
> 
> Paul