Re: [Gusdev-gusdev] improving GUS install/evolution

[Steve F. <sfi...@pc...>]

Terry-

Great to hear that your paper will be in Applied Bioinformatics.

I just had a visit to UGA where Jessica Kissinger's lab is using GUS.

We have been discussing the GBParser and the parser of TIGR XML, 
considering whether they should be re-written/unified.  Ed Robinson 
there is focusing on this.

The approach we have been thinking of is to have both of them parse into 
a common object model (either bioperl, if it turns out they have an 
appropriate model, or the existing and possibly modified CBIL genbank 
objects).   They would both share a layer that writes the object model 
into GUS objects and submits.   They would differ in the layer that 
parses into the model. 

Ed has been thinking about the layer to parse into the objects and has 
mentioned a declarative mapping from the source into the objects.

I told him about your work, so, this is perfect timing.

Can we get a pre-print of your paper?

Also, you mention below that you are not using plugins, and are 
transforming directly from XML into the schema.   Sounds like you are 
not using the object layer.   But, the object layer and ga have benefits 
like handling overhead rows, algorithm tracking and versioning.   How do 
you handle that stuff?   Would it be possible to modify your approach to 
fit into the plugin framework so that, instead of going to the schema 
directly, it goes into the objects?   Or, am did i misinterpret and are 
you doing this already?

Thanks,
steve

Terry Clark wrote:

>One component of schema evolution is getting data 
>into the schema; another side is moving new data 
>into existing schema. 
>
>Folks at UChicago and I have worked out an XML
>based strategy for incorporating data into GUS. 
>The idea is around a GUS XML that contains key constraints
>as attributes, which is interpretted by a byacc
>parser and processor (together these are a plugin).
>The data-to-schema routing happens by way of the XML
>and grammar.
>
>There is a paper to appear in Applied Bioinformatics
>(probably the next issue) about our GUX XML processed
>by XMLGUS. We find this a useful tool in our production 
>work. 
>
>Now writing plugins is unnecessary for us apart 
>from occasional special cases, which XMLGUS has hooks for.
>We now use XML and context free grammars to define the
>mapping from data to schema, in most of our cases. 
>There is one case my colleague found it easier to write
>a plugin from scratch    - I have yet to study this to
>see what caused this.
>
>The GUS install needs a way to incorporate local tables.
>A GUS table need not be part of the canonical set of 
>tables; some can be site specific. We add local tables now, 
>but the process is very ad hoc. 
>
>Presently, we are working now on automating the production
>of the XMLGUS framework, and the table problem. It would
>be useful to see what people are thinking about and
>where the needs are.
>
>Maybe such ideas could be discussed at a GUS meeting?
>
>cheers,
>
>Terry
>
>
>
>
>
>On  0, Sucheta Tripathy <su...@vb...> wrote:
>  
>
>>Steve,
>>
>>We at VBI welcome the idea of releasing a version with install data.
>>
>>This has been the main cause of concern for GUS users who had to spend
>>lots of time finding the pre-requisite data .
>>
>>We can certainly contribute to the initial dataset.
>>
>>In addition we would be more interested to initiate a discussion on
>>functional annotation.
>>
>>cheers
>>
>>Sucheta
>>
>>    
>>
>>>folks-
>>>
>>>some people here at cbil and also at uga have begun thinking about how
>>>to improve the GUS install, and how to manage schema evolution.
>>>
>>>the two pieces of the gus install are:
>>>   (1) the database schema and initial data
>>>   (2) the software
>>>
>>>mike at cbil has made progress on (1), and it seems like alan, ed and
>>>haiming at uga have ideas about this.
>>>
>>>mike is promising to write to the group to give a report on his efforts.
>>>
>>>so, maybe that will be a good way to get the discussion rolling....
>>>
>>>if folks have ideas or energy to tackle (2) that would be great.
>>>
>>>steve
>>>
>>>
>>>
>>>
>>>
>>>-------------------------------------------------------
>>>This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
>>>Use IT products in your business? Tell us what you think of them. Give us
>>>Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out
>>>more
>>>http://productguide.itmanagersjournal.com/guidepromo.tmpl
>>>_______________________________________________
>>>Gusdev-gusdev mailing list
>>>Gus...@li...
>>>https://lists.sourceforge.net/lists/listinfo/gusdev-gusdev
>>>
>>>      
>>>
>>-- 
>>Sucheta Tripathy
>>Virginia Bioinformatics Institute Phase-I
>>Washington street.
>>Virginia Tech.
>>Blacksburg,VA 24061-0447
>>phone:(540)231-8138
>>Fax:  (540) 231-2606
>>
>>
>>-------------------------------------------------------
>>This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
>>Use IT products in your business? Tell us what you think of them. Give us
>>Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
>>http://productguide.itmanagersjournal.com/guidepromo.tmpl
>>_______________________________________________
>>Gusdev-gusdev mailing list
>>Gus...@li...
>>https://lists.sourceforge.net/lists/listinfo/gusdev-gusdev
>>    
>>
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
>Use IT products in your business? Tell us what you think of them. Give us
>Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
>http://productguide.itmanagersjournal.com/guidepromo.tmpl
>_______________________________________________
>Gusdev-gusdev mailing list
>Gus...@li...
>https://lists.sourceforge.net/lists/listinfo/gusdev-gusdev
>  
>