Semantic resources project/Meeting notes/2009-09-10

= Notes from September 10 meeting =

Taken by: Kaitlin Thaney

Attendees: Jonathan Rees, Paolo Ciccarese, Kaitlin Thaney, Tim Danford, Sudeshna Das

I. Antibodies


 * Paolo summarized definitions of antibodies - did analysis of 4 companies to get idea of what we can grab from them. Will circulate to TimD and post in Google group

'''II. Use cases'''


 * What is the migration path to get into one of the communities? (become part of the codebase)


 * SD: Deal more with editors of sites than users, so for use case development, should be in contact with editors as a start: Lisa Gerard (Stembook), Mike Rogan (PDOnline)


 * SD to make introduction. TimD to conduct interviews and information gathering.

'''III. NC install'''


 * Paolo: What is in the knowledgebase now or does it include what we are working on now


 * JAR: How is the path to integrate that into SCF to get to final users /What is the migration path?

In terms of looking at the development cycle, one has idea for use case -> development of resource -> creation of UI -> become part of SCF.

What do you do next socially and technically to interface with the projects? Is this something the SCF team will work on actively, or after the project?


 * SD: Software apsect - have a way that if a resource has a URI, can make an instance and integrate in to the SCF system. There's a way to do SPARQL queries, but not optimal. In terms of interaction with user community, usually don't deal with users of sites but editors of the site.


 * JAR: That will come out as we write the use cases ... who are the users? Are we talking just about editors or as individual researchers? Decision we can make.


 * Paolo: Once we have the antibody repository, how will it emerge in the SCF communities?


 * SD: Within a month, Stefan and Sam (new hires at SCF / PDOnline) will be up to speed to integrate


 * Paolo: How to integrate, though?


 * SD: For antibodies - have articles in Stembook, want to find antibodies articles used there, link to antibody resource. Another thing of interest, to hook up with SC MTA agreements (Coriell?)

Discussion of hardware install for NeuroCommons instance


 * Set up on the Amazon cloud?


 * Memory needed: 32 gigabytes for all, can slim down to 4 or 8.


 * JAR: there would be some development phase and experimentation as to what goes into the triple store. At some point, decide it's good enough for SCF or publicly licensed distribution.

But if something has to be kept closed or not stable to release, keep as special edition in triple store.

Components: server, triple store, knowledgebase, modules wanted to add to the triple store / distribution.


 * SD: thinks appropriate to have on the cloud for our rapid development and testing.


 * SD: sending email to Marco, JAR, TimD. TimD to be charged with that task.

'''IV. Next work


 * TimD - install on the Cloud on SCF side.
 * TimD - sync up with JAR/alan - conduct interviews with people in the lab, Stembook, PDOnline, faculty people at MIND (institute for neurodegenerative disease)

Questions: who the editors / users are, how an antibody resource would be useful, also other resources (maybe add pathways?) what resources?


 * TimD: Another resource: metadata associated with experiments that's related to resources that already exist. just being able to import the metadata for the experiements and relate

Example: for a binding experiment - what protein measuring, and what measurement is being used.


 * TimD to write up resource for review.

Adjourned