Re: [Gmod-ajax] Article: Using game design to build the next Digg or Flickr

[Chris M. <cj...@fr...>]

On Aug 24, 2007, at 5:41 PM, Mitch Skinner wrote:

> Ian Holmes wrote:
>> How would this compare with existing systems e.g. Ensembl MART?
>>
>
> I've only spent a bit of time with it, but I don't see a way to  
> view the
> results of a biomart query in the genome browser, which is what I was
> thinking.  In addition to creating new tracks by uploading, we could
> allow people to create tracks by defining queries.  It should be easy
> for us to track how popular these are, so there's a natural connection
> to the points and feedback/recognition items from the game/webapp  
> article.
>
> Also, while you can specify a list of IDs to biomart, that requires  
> the
> user to manage their own collection.  We could certainly make  
> collection
> management a lot easier (e.g., an "add this to my collection"  
> button on
> feature popups).  Centralizing collections also makes them easier to
> share (like the "post your hotlist" thing, with individual features
> instead of tracks).

I like this idea, but in the interests of not seeing us diverge too  
far from the core plan I'd like to place this in the context of a  
wider architecture.

I think GBrowse-ajax should make the assumption that we will soon  
live in an ideal bioinformatics nation where data sources will expose  
powerful query interfaces - MART, InterMine, or more ontology-based  
schemes - through some appropriate HTTP API or RESTful architecture.

GBrowse should provide a way of slurping a bunch of feature IDs and  
showing them in an appropriate way. Or, conversely, a way of  
exporting a bunch of IDs, perhaps from a ROI.

So long as gbrowse is mashup/plugin friendly then it should be  
possible to wrap this functionality in the way you describe.

> In addition to having a nice query building UI like biomart does, I
> think it makes sense to expose something lower level, like (say) raw
> SPARQL.  Most biologists won't use it, but if we make it easy for more
> advanced users to share their queries with other people then it's  
> still
> pretty useful IMO.

SPARQL is fairly limited. However, it is the standard query language  
for RDF.

You can assume that a SPARQL wrapper for Chado will be available  
sometime in the next half of year or so. We've already done this for  
the GO database[1]. Again, I think GBrowse should just worry about  
accepting lists of IDs and/or regions - this would be the main point  
of contact. Some enterprising semweb fan will then hopefully cobble  
together a non-horrible SPARQL interface that pipes the results into  
a GBrowse display.

> But these are all just random ideas.  Mainly, I wanted to ask for help
> translating game ideas into stuff that relates more concretely to what
> we're doing.
>
> Mitch
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Gmod-ajax mailing list
> Gmo...@li...
> https://lists.sourceforge.net/lists/listinfo/gmod-ajax
>