[BioC] Biomart with human genes NCBI 35/hg17

[Yan Jiao]

Hi Steffen,

I also tried using lift genome annotations for mapping different built: http://genome.ucsc.edu/cgi-bin/hgLiftOver

It helps a bit, but some regions it couldn't give me the mapping result, e.g.
Failed input regions:

    #Split in new
    chr1:0-245500000
    #Partially deleted in new
    chr21:0-46900000

I suppose the region is too long, any suggestions?

Many thanks

Yan




-----Original Message-----
From: Steffen Durinck [mailto:durinck.steffen at gene.com] 
Sent: 26 July 2011 15:52
To: Yan Jiao
Cc: bioconductor at r-project.org
Subject: Re: [BioC] Biomart with human genes NCBI 35/hg17

Hi Yan,

Ensembl 44 is to far back and the BioMart interface has changed
significanlty since, as a result you can't query such old archive.
Any Ensembl version from Ensembl 52 and upwards should be fine though,
unfortunately this won't help you as you want NCBI35, and Ensembl 52
is NCBI 36.

Note that a better way to access the archives is to specifiy the host
url as described in the vignette:

ensembl54=useMart(host='may2009.archive.ensembl.org',
biomart='ENSEMBL_MART_ENSEMBL', dataset='hsapiens_gene_ensembl')


Cheers,
Steffen

On Tue, Jul 26, 2011 at 7:02 AM, Yan Jiao <y.jiao at ucl.ac.uk> wrote:
> Dear All,
>
>
> I'm trying to use the previous built (within Biomart)for human genes to get gene location, It seems like Ensembl 44 is the version I should use for NCBI 35/hg17, I did following:
>
> ensembl =useMart("ensembl_mart_44",archive=TRUE)
> listDatasets(ensembl)
> [1] dataset     description version
> <0 rows> (or 0-length row.names)
>
> So you could see there is no dataset available,
> So if I do:
>
> ensembl =useMart("ensembl_mart_44",dataset="hsapiens_gene_ensembl",archive=TRUE)
> I got:
> Error in useDataset(mart = mart, dataset = dataset, verbose = verbose) :
>  The given dataset:  hsapiens_gene_ensembl , is not valid.  Correct dataset names can be obtained with the listDatasets function.
>
> has anyone come across the same problem?
>
> Many thanks
>
> Yan
>
>
>        [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
>