From gregorysbrown at mac.com  Mon Nov  5 16:16:07 2007
From: gregorysbrown at mac.com (Greg Brown)
Date: Mon, 5 Nov 2007 13:16:07 -0800
Subject: [Gramene-extdev] Question about the 'NULL' protein domain in
	Arabidopsis
Message-ID: <8785C3D8-9DE2-41B0-81B3-329A344DD794@mac.com>

Hi,

We first found this 'NULL' domain type in our mirrored version 26.  
There are two things we see that is strange.

1. There are dozens of Arabidopsis genes associated with the 'NULL'  
protein domain that according to InterPro is not a valid domain.  
http://www.gramene.org/Arabidopsis_thaliana/domainview?domainentry=NULL

2. If one of the peptides is selected with a 'NULL' domain the  
protview page lists what apppears to be ALL protein domains, not just  
the ones assigned to the particular protein. Here is an example;  
http://www.gramene.org/Arabidopsis_thaliana/protview? 
transcript=AT1G02405.1-TAIR;db=core

We assume there is some small systematic error with assignments of  
protein domains. Do you anticipate a patched dataset for this version  
or will you wait until version 27 to fix it? Thanks

Greg Brown
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://brie4.cshl.edu/pipermail/gramene-extdev/attachments/20071105/0717559b/attachment.html>

From whs at ebi.ac.uk  Mon Nov 12 16:49:34 2007
From: whs at ebi.ac.uk (Will Spooner)
Date: Mon, 12 Nov 2007 21:49:34 +0000 (GMT)
Subject: [Gramene-extdev] Question about the 'NULL' protein domain in
	Arabidopsis
In-Reply-To: <8785C3D8-9DE2-41B0-81B3-329A344DD794@mac.com>
References: <8785C3D8-9DE2-41B0-81B3-329A344DD794@mac.com>
Message-ID: <Pine.LNX.4.64.0711122141160.22633@parrot.ebi.ac.uk>

Hi Greg,

Thanks for the feedback. We get the arabidopsis data the data directly 
in an ensembl database from NASC, so I will check back with them about 
these domains. It is unlikely that this will be fixed until NASC release a 
new database, unless we run the gramene pipeline on arabidopsis.

All the best,

Will

On Mon, 5 Nov 2007, Greg Brown wrote:

> Hi,
>
> We first found this 'NULL' domain type in our mirrored version 26. There are 
> two things we see that is strange.
>
> 1. There are dozens of Arabidopsis genes associated with the 'NULL' protein 
> domain that according to InterPro is not a valid domain. 
> http://www.gramene.org/Arabidopsis_thaliana/domainview?domainentry=NULL
>
> 2. If one of the peptides is selected with a 'NULL' domain the protview page 
> lists what apppears to be ALL protein domains, not just the ones assigned to 
> the particular protein. Here is an example; 
>
http://www.gramene.org/Arabidopsis_thaliana/
protview?transcript=AT1G02405.1-TAIR;db=core
>
> We assume there is some small systematic error with assignments of protein 
> domains. Do you anticipate a patched dataset for this version or will you 
> wait until version 27 to fix it? Thanks
>
> Greg Brown