[ensembl-dev] Biomart Genomic Coding Start/End

[Simon Chan]

Hello,

I'm using BioMart to fetch all protein coding portions of all exons of all transcripts of my favourite gene (Ensembl Genes 67, Homo sapiens genes GRCh37.p7).  In the past, 'Genomic coding start' and 'Genomic coding end' would give me all exon coordinates minus UTRs.  However, now it seems to return the UTR coordinates as well.  Perhaps this is related to the UTR issue reported by Beat below?

Thank you.
Simon.

From: dev-bounces at ensembl.org [mailto:dev-bounces at ensembl.org] On Behalf Of Rhoda Kinsella
Sent: Tuesday, May 22, 2012 3:38 AM
To: Ensembl developers list
Subject: Re: [ensembl-dev] Ensembl release 67 is out!

Hi Beat
We are currently looking into why this has occurred but it is certainly a bug in our mart databases. I will respond as soon as we have a fix in place. Thank you for reporting this and apologies for any inconvenience.
Regards
Rhoda

On 19 May 2012, at 15:55, Beat Wolf wrote:


Hi,


i'm a computer scientist and not a biologist, so perhaps i missed something about the DMD gene changing recently.
I have an automated unit test for my application that uses biomart. It happens to test the DMD-001 transcript of the DMD gene.
In the 66 version of enembl, the UTR3 coordinates of the exon number 79 where:


31137345

to 31140035

which was correct. Now with the version 67 of ensembl, the UTR3 region in the same exon is:


31137345 to

31137344

Which can not be correct, because that means that UTR3 has a length of 1. The other numbers don't match up either.


Did the DMD gene change? Is the information stored differently in ensembl? What values do i need to reconstruct the true UTR3 region? Or is this a bug in the new ensembl version. If yes, how many other genes are affected? Currently i will stay on the 66 version thanks to the archives, but i have some connection issues with it, it seems to be much less stable than the 67 version.


Greetings


Beat Wolf


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20120523/1a96d498/attachment.html>