[BioC] biomaRt : multiple entrezgene ids to one affy id?

[Sean Davis]

On Monday 02 April 2007 11:19, Weiwei Shi wrote:
> Dear all:
>
> Here is some piece of output from my query:
> > mart=useMart(biomart="ensembl", dataset="hsapiens_gene_ensembl")
>
> Checking attributes and filters ... ok
>
> > getBM(attributes=c("affy_hg_u95a", "entrezgene","ensembl_transcript_id"),
> > filters="affy_hg_u95a", values="31330_at", mart=mart)
>
>   affy_hg_u95a entrezgene ensembl_transcript_id
> 1     31330_at       6223       ENST00000221975
> 2     31330_at     728953       ENST00000221975
> 3     31330_at     731572       ENST00000221975
>
> I do not understand why there are multiple entrezgene ids
> corresponding to one affy id. Did I miss something here?

Unfortunately, this is what happens when you start mapping between identifiers 
defined by different authorities.  If you look at the entry for ensembl 
transcript, you will see that it maps to three different EntrezGene IDs.  
Looking a bit further, you will notice that only one of them (6223) is based 
on a mapping between a non-predicted sequence.  The other two are predicted 
sequences.  

Here is the entry in ensembl for reference:

http://www.ensembl.org/Homo_sapiens/geneview?gene=ENSG00000105372

Sean