[open-science] NeuroStars, a Q&A site for neuroscience

Roman Valls Guimera roman at incf.org
Mon Jul 7 08:40:45 UTC 2014


Hi Bill,

Glad and excited to hear you guys at MozillaScience are interested in NeuroStars and *-stars! :)

Istvan Albert, the original author of BioStars and his team are working on a Bittorrent tracker for the backend of BioStars. I’m sure Istvan can fill you in with more details if need be:

https://github.com/ialbert/biostar-central/commit/5b6401e7576a6f21

When it comes to NeuroStars and INCF, there’s an ongoing effort to support iRODS (integrated Rule-Oriented Data System)[1] data storage platform, used already by many research institutions on completely disparate fields (neuro, oceanographics, genomics, ...).

We are facing **many** development issues right now, such as the usability of the iRODS UI’s (both desktop and web)[7]. Here you have some discussions that have been sparked very recently with one of the core iRODS developers, Mike Conway, also on CC now:

https://github.com/DICE-UNC/idrop/issues/30
https://github.com/DICE-UNC/idrop/issues/31

More specifically, as part of the threads above, you can see an ongoing collaborative mockup of the new irods-web interface:

https://gomockingbird.com/mockingbird/#jm4tvi7/QRhyJN


Systems like git-annex[2] and its assistant[3], in use to share interesting neuroscientific data such as "Study Forrest” [4] by Mikael Hanke, could be also integrated in *-stars engine.

Last but not least, to bring some good provenance and semantic value to this whole mashup, NIDASH[6] (Satra and Nolan on CC) is an excellent candidate.

To sum up a bit, I think that having those data sharing backends easily deployable[5] (by anyone!) through a free Q&A system such as *-stars has a lot of potential as an integrating hub for scientific data.



About your question on the (binary) nature of neuro datasets, you have a good sample if you browse study forrest:

http://psydata.ovgu.de/forrest_gump/


Best regards,
Roman


[1] http://irods.org/
[2] http://git-annex.branchable.com/
[3] http://git-annex.branchable.com/assistant/
[4] http://studyforrest.org/
[5] https://github.com/nimiq/ansible-biostar
[6] http://nidm.nidash.org/
[7] https://github.com/DICE-UNC/idrop/wiki/iDrop-Installers

7 jul 2014 kl. 00:32 skrev Bill Mills <mills.wj at gmail.com>:

> Hi Roman,
> 
> Alright - now that our project selection phase has completed, we would definitely like to find a way to move forward with your goals for Neurostars!  Two things come to mind:
> 
> - Your list of GitHub issues look great; I think they're concrete enough that designers and web developers could jump right in.
> - In your original email, you mentioned being interested in data sharing utilities; can you tell us a bit more about your ideas and goals in this space?  Both Mozilla and I are very keen to funnel developers towards helping build open data sharing solutions, so let us know a bit more about that goal and maybe we can make some progress there.  One place to start, is helping us get a grasp of just what kind data you want to share - text files?  Binaries?  Images?  Let us know!
> 
> We are definitely going to move forward looking for support for Neurostars; let's see if we can bring the data sharing goal into focus, and go from there!
> 
> 
> On Fri, Jun 27, 2014 at 4:31 PM, Kaitlin Thaney <kaitlin at mozillafoundation.org> wrote:
> Thanks, both! Will be sure to give this a look. :)
> 
> - K
> 
> sent on the move. apologies for typos.
> 
> 
> > On 28 Jun 2014, at 08:21, Roman Valls Guimera <roman at incf.org> wrote:
> >
> > Hi Bill,
> >
> > Thanks for your feedback!
> >
> > This afternoon Satra and I sat down and put together some, more specific, set of the issues over GitHub:
> >
> > https://github.com/incf/biostar-central/issues?labels=MozillaScience&state=open
> >
> > Please feel free to ask further either over GitHub or this mailing list.
> >
> > Hope that helps!
> > Roman
> >
> >
> >> 27 jun 2014 kl. 08:41 skrev Bill Mills <mills.wj at gmail.com>:
> >>
> >> Hi Roman,
> >>
> >> Neurostars looks like a great tool with demonstrated potential, and the things you're thinking about getting developer help on sound like potential good fits for our project.  We'll be choosing projects to move ahead with around the end of next week, but before then and as part of the decision process, I'd like to zero in a bit on exactly the tasks you'd like developers to focus on.  Points 1-3 all sound interesting, but pretty broad; any one of those broken down into a collection of smaller goals could be an excellent project at the right scale.  Also, the UI & django work you mention could be real watersheds here, since as you point out there is low to no domain knowledge getting in the way of onboarding designers and developers there.  If you can choose the goal you most want outside help on and break it down one more zoom level, I think we'll have a strong collaboration proposal on a valuable project.  Let us know, looking forward to it!
> >>
> >>
> >> On Thu, Jun 26, 2014 at 7:31 AM, Roman Valls Guimera <roman at incf.org> wrote:
> >> Dear Kaitlin,
> >>
> >> We would like to know whether a project that we have been working on is along the lines of what you need for the call for #mozscience.
> >>
> >> Briefly, http://neurostars.org is a question and answer site for discussing and sharing knowledge about neuroscience and neuroinformatics that is based on http://biostars.org, a highly successful question and answer site for the Bioinformatics field.
> >>
> >> NeuroStars.org was forked to foster a cultural shift in the biomedical imaging community away from software-specific mailing lists to a more efficient platform for collaborative problem solving like stackoverflow.com (which is not open source), since many of the posts on the software-specific mailing lists were identical in either content or concept.
> >>
> >> Currently, Paolo, our Google Summer of Code 2014 student is taking care of pushing new and exciting features to the site:
> >>
> >> http://nimiq.github.io/my-summer-of-code/
> >>
> >> We pullrequest back to the original author, Istvan Albert, so that the bioinformatics community (among other sci communities) can benefit from our additions:
> >>
> >> https://github.com/ialbert/biostar-central
> >>
> >> Our current broad development goals are to:
> >>
> >> 1. Provide an RDF integration that is able to cross reference resources within and across sites and neuroinformatics resources such as NIDASH, a scientific provenance standard for bioimaging data.
> >> 2. Integrate and expose different dataset sharing protocols such as Bittorrent, git-annex, and iRODS, to simplify data sharing within science.
> >> 3. Integrate ORCID, a popular researcher identifier to increase publication awareness among scientists working on related fields.
> >>
> >> As you might notice when visiting the site, there’s room for improvement, both UI and backend. We could definitely benefit from good graphics designers and javascript experts, for instance.
> >>
> >> At the end of the day, *stars sites are Django apps, so the domain-specificity is rather low, allowing developers to jump in rather quickly. On the other hand we definitely benefit and learn from scientific needs and feedback from experts, on CC and also on GitHub:
> >>
> >> https://github.com/INCF/biostar-central/issues/36
> >>
> >> I hope it sounds interesting!
> >>
> >> Best regards,
> >> Roman
> >>
> >>
> >>
> >> --
> >> Best Regards,
> >> Dr. Bill Mills
> >> TRIUMF
> >
> 
> 
> 
> -- 
> Best Regards,
> Dr. Bill Mills
> TRIUMF



More information about the open-science mailing list