[DAS] [Gmod-gbrowse] Adding comments to a gbrowse database

Tue Nov 2 13:07:15 UTC 2010

Hi Dan,

Now I see what you mean - basically the wiki and DAS source have the same data/comments, and there are two ways to enter it (writeback inside the client OR the wiki itself). This sounds good and certainly feasible. I guess there will be details about the format the comments take in the Wiki, but I'm sure it would work. I agree that the great thing about an approach like this (and DAS writeback's "independent" architecture) is the open model. This, and the fact that it would also be possible to have the comments propagated into multiple clients (the usual DAS distributed approach) make things much more appealing than just a single isolated database of user comments.

It's probably worth looking at the provenance/users aspect of the wiki integration to see how this can work with writeback. Gustavo can say more about how writeback handles user tracking I expect, but authentication is often the most difficult thing when it comes to implementing these things, I find. Specifically in this case: "when I edit something in the client, can people see who made the edit on the wiki itself, and if so is the process by which that happens secure?"

I definitely encourage you to look further into it and am keen to find out how you get on.

Cheers,
Andy

On 2 Nov 2010, at 10:14, Dan Bolser wrote:

> Here is how I thought it would work (please let me know if this sounds
> about right, or if I made some mistakes in my understanding of the way
> DAS writeback will work):
> 
> 1) I was thinking that the ProServer DAS server should (or at least
> could) implement the DAS writeback extension.
> 
> 2) The genome browser client (or some other client) would talk to the
> DAS server using the DAS writeback protocol once a comment was added
> (i.e. I'm relying on a sensible client that can 'properly' handle
> comments).
> 
> 3) I expect that ProServer will offload the job of actually storing
> the comment on a datasource adaptor.
> 
> 4) I thought I'd extend the MediaWiki adaptor that I wrote for reading
> feature data from the wiki to also write feature data there too.
> (Comments are just features right?)
> 
> 5) You could then use either the genome browser client (or some other
> client) or the wiki directly to add and/or edit comments. i.e. the
> wiki is just another feature editing client / DAS datasource.
> 
> 
> I like this solution because I think 'community annotation' should be
> in the 'community', not on some server behind closed doors somewhere
> (somehow a wiki 'feels' more 'open' to me than some bespoke database
> x). There are plenty of wikis that could be used as back ends, and
> 'wiki' is a good way to do community annotation. They handle issues
> related to provenance, QC, versioning, authorship, motivation,
> publicising, publication, etc., etc. Again, MediaWiki is a good
> choice, because their is a lot of development for that platform, and I
> can easily imagine tagging, 'micro-publication' and syndication
> extensions being leveraged for this kind of data.
> 
> 
> Does it sound reasonable?
> 
> 
> Cheers,
> Dan.
> 
> BioWiki conference:
> http://www.nettab.org/2010/
> 
> 
> On 30 October 2010 12:41, Andy Jenkinson <andy.jenkinson at ebi.ac.uk> wrote:
>> Hi Dan,
>> 
>> Sounds interesting. How do you see that working?
>> 
>> Cheers,
>> Andy
>> 
>> On 29 Oct 2010, at 16:28, Dan Bolser wrote:
>> 
>>> My thinking was that the DASWiki approach could also support the DAS
>>> writeback extension.
>>> 
>>> On 29 October 2010 11:58, Gustavo Adolfo Salazar Orejuela
>>> <gsalazar at cs.uct.ac.za> wrote:
>>>> Hello There,
>>>> 
>>>> I have been working in an extension to DAS1.6 called
>>>> writeback(http://www.biodas.org/wiki/DAS1.6E#DAS_writeback), this will allow
>>>> to create new versions of an existing feature. So, putting new notes into a
>>>> feature is just a creation of a new version of the feature. The server
>>>> implementation of the writeback extends MyDas
>>>> (http://code.google.com/p/mydas/) and you can get a mydas+writebackk
>>>> datasource by SVN (https://mydas.googlecode.com/svn/writeback-datasource).
>>>> We are close to release the plugin for Dasty3 to use the writeback server,
>>>> however this first client is a protein client, and so far there is not a
>>>> development for a genome client with DAS writeback capability, but I am
>>>> quite interested in this.
>>>> I hope to be able to announce the release of the writeback client in Dasty3
>>>> in a few days, so you will be able to see it working and see if this
>>>> approach goes in the same direction of what you are looking for.
>>>> 
>>>> Regards,
>>>> 
>>>> Gustavo.
>>>> 
>>>> 2010/10/29 Dan Bolser <dan.bolser at gmail.com>
>>>>> 
>>>>> Hi Bob,
>>>>> 
>>>>> This is a nice idea. In my spare time I've been working on running a
>>>>> DAS server from a MediaWiki data source. (MediaWiki is the wiki that
>>>>> runs Wikipedia).
>>>>> 
>>>>> In theory, you could have a comments track that was pulled in from the
>>>>> 'DASWiki', and clicking the details link would take you to the wiki
>>>>> page to edit the comment. Going back to the GBrowse, you would then
>>>>> see the updated data from the wiki page. It wouldn't be too hard to
>>>>> have a similar workflow for adding new comments.
>>>>> 
>>>>> I saw a talk from some people at KeyGene who are doing something
>>>>> similar (loading DAS into a wiki and linking it to GBrowse to generate
>>>>> query reports for specific regions). You can see the talk here:
>>>>> 
>>>>> SMWCon Fall 2010 / Sunday / Rudi van Bavel
>>>>> * http://semantic-mediawiki.org/wiki/SMWCon_Fall_2010#Program
>>>>> ** https://docs.google.com/present/view?id=dg56kh6m_344gh2dg72g
>>>>> **
>>>>> http://srv-hrl-32.web.pwo.ou.nl//informatica/SMWCon/Van%20Bavel%20bulk%20loading.wmv
>>>>> 
>>>>> 
>>>>> If anyone wants to take what I have done so far, my (sadly neglected)
>>>>> prototype is here:
>>>>> 
>>>>> http://das.referata.com/wiki/Main_Page
>>>>> 
>>>>> 
>>>>> Kick me a few times and I'll get the underlying code into shape. A
>>>>> year or two back I did have a working prototype based on a MediaWiki
>>>>> ProServer adaptor.
>>>>> 
>>>>> 
>>>>> If you are interested in community annotation should think about attending
>>>>> this:
>>>>> 
>>>>> Joint NETTAB 2010 and BBCC 2010 workshops focused on Biological Wikis:
>>>>> http://www.nettab.org/2010/
>>>>> 
>>>>> 
>>>>> Cheers,
>>>>> Dan.
>>>>> 
>>>>> On 26 October 2010 20:31, Bob Muller <bmuller at stanford.edu> wrote:
>>>>>> We are working with a small research project that doesn't have funds for
>>>>>> a full-scale database for features, so they want to know:
>>>>>> 
>>>>>> How difficult would it be to add a comment section editable by users to
>>>>>> gbrowse?  When you click on a feature in gbrowse it shows a detail page for
>>>>>> that feature, is there any way we could enable edits that could be written
>>>>>> back to the underlying mysql database that gbrowse runs on?
>>>>>> 
>>>>>> Their idea is to use the SeqFeature::Store database to hold "comments"
>>>>>> entered through GBrowse as a way to annotate their genome. I'd be interested
>>>>>> in whether this is possible in GBrowse 1.7 or in 2.x.
>>>>>> 
>>>>>>   --Bob Muller, TAIR (www.arabidopsis.org)
>>>>>> 
>>>>>> 
>>>>>> ------------------------------------------------------------------------------
>>>>>> Nokia and AT&T present the 2010 Calling All Innovators-North America
>>>>>> contest
>>>>>> Create new apps & games for the Nokia N8 for consumers in  U.S. and
>>>>>> Canada
>>>>>> $10 million total in prizes - $4M cash, 500 devices, nearly $6M in
>>>>>> marketing
>>>>>> Develop with Nokia Qt SDK, Web Runtime, or Java and Publish to Ovi Store
>>>>>> http://p.sf.net/sfu/nokia-dev2dev
>>>>>> _______________________________________________
>>>>>> Gmod-gbrowse mailing list
>>>>>> Gmod-gbrowse at lists.sourceforge.net
>>>>>> https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>>>>>> 
>>>>> 
>>>>> _______________________________________________
>>>>> DAS mailing list
>>>>> DAS at lists.open-bio.org
>>>>> http://lists.open-bio.org/mailman/listinfo/das
>>>> 
>>>> 
>>>> 
>>>> --
>>>> Gustavo Adolfo Salazar Orejuela
>>>> 
>>> 
>>> _______________________________________________
>>> DAS mailing list
>>> DAS at lists.open-bio.org
>>> http://lists.open-bio.org/mailman/listinfo/das
>> 
>> 