Re: [GUSDEV] Gene Association data load (GFUS 3.0)

[Jonathan S. <js...@pc...>]

Sanjeev:

The table dots.GoAssociationInstanceLoe is a 'controlled' vocabulary  
of that indicate the different lines of evidence one might use when  
assigning a GO term to an entity.  We developed it because we load GO  
term associations from multiple sources.  For example, the PlasmoDB  
project got GO annotations from the sequencing centers and we also  
made our own GO function term predictions.  The link in  
dots.GoAssociationInstance to this table allows us to distinguish the  
two.

I would use GUS::Common::Plugin::UpdateGusFromXML to load  
dots.GoAssociationInstanceLoe.  Here's an example of the XML you  
might use as input to that plugin:

<DoTS.GOAssociationInstanceLOE>
   <name>Human Annotators</name>
   <description>Members of a team of experts look at the gene name  
and think up a GO term.</description>
</DoTS.GOAssociationInstanceLOE>

<DoTS.GOAssociationInstanceLOE>
   <name>Automatic Annotation</name>
   <description>A very smart programs looks at the DNA sequence and  
thinks up a GO term.</description>
</DoTS.GOAssociationInstanceLOE>

This bit of XML defines two source of annotation; you can define more  
by running the plugin again or by putting more in the file.  Note  
that dots.GoAssociationInstance also has an  
external_database_release_id column which can pin-point the source of  
the data very precisely.  The goal of dots.GoAssociationInstanceLoe  
is a broad classification of sources.  Remember there is also the  
evidence code which also indicates how the association was formed.   
If you don't think you'll care about categorizing sources in this  
way, just enter a single 'some method' in  
dots.GoAssociationInstanceLoe and use that for all association sources.

Hope that helped.

Jonathan

Jonathan Schug, Ph.D.

Center for Bioinformatics
Computational Biology and Informatics Lab
University of Pennsylvania

js...@pc...
(215) 573-3113 voice
(215) 573-3111 fax

1413 Blockley Hall, Philadelphia, PA 19014-6021



On Sep 7, 2005, at 10:51 PM, Chris Stoeckert wrote:

> Hi Sanjeev,
> These tables were created for our GO prediction pipeline (see  
> http://www.cbil.upenn.edu/GO/). You might want to check if this is  
> really appropriate to what you want to do.
> The person responsible for this (Dave Barkan) has left the group.  
> Jonathan, Debbie any insights to provide?
> Thanks,
> Chris
>
> On Sep 6, 2005, at 10:22 AM, Kumar, Sanjeev (Contr) wrote:
>
>
>> Hi,
>>   I am trying to load GOAssociation data by usingga
>> GUS::GOPredict::Plugin::LoadGoAssoc
>>   plug in in GUS3.0 .
>>   I am getting an error at the initial stage of data load. It is  
>> looking
>> for data in
>>   table called "Dots.GOASSOCIATIONINSTANCELOE" , because
>> "Dots.GOASSOCIATIONINSTANCE" has
>>   foreign key relationship with the first table.
>>   I don't know how I am going to get data in the
>> "Dots.GOASSOCIATIONINSTANCELOE" table.
>>
>>   Can some one help me?
>>
>> Thanks
>> Sanjeev
>>
>>
>> -------------------------------------------------------
>> SF.Net email is Sponsored by the Better Software Conference & EXPO
>> September 19-22, 2005 * San Francisco, CA * Development Lifecycle  
>> Practices
>> Agile & Plan-Driven Development * Managing Projects & Teams *  
>> Testing & QA
>> Security * Process Improvement & Measurement * http://www.sqe.com/ 
>> bsce5sf
>> _______________________________________________
>> Gusdev-gusdev mailing list
>> Gus...@li...
>> https://lists.sourceforge.net/lists/listinfo/gusdev-gusdev
>>
>