SDIWG:Meeting Minutes 20051110

Tcon with new Centers: On phone/breeze: Brian Athey, Suzanna Lewis, Karen Skinner, Don Jenkins, Chris Bliton, Walter Meixner, Aris Floratos, Dan Gallahan, Daniel Rubin (Exec Dir), Barry Smith (Ongology dissemination).

Note Taker: Peter Lyster (if I remember correctly)

Karen: Introduces agenda and members of the three new teams.

Peter: The agenda is: (1) get a handle on the new NCBCs’ own self-organizing activities and (2) fine tune the agenda for the December meeting—perhaps (1) will feed into (2). I also want to alert you to ongoing activities that are organized by the Software and Data Integration Working Group (SDIWG), so it might be useful to bring you up to speed on activities related to: (a) Yellow pages, (b) software knowledgebase, (c) collaborative software development environment.

ONGOING ACTIVITIES OF 3 GROUPS: Brian: NCIBI works with Musen group. History of DARPA effort. Molecular/SNPs/GO. Also similarities of Columbia with Ann Arbor. Jill Mesirov at Broad. Taverna workflows. There are synergies—hence Brian thinks it might be useful, and can help with ‘big P’ ref J. Whitmarsh. Brian and Suzanne was on SDIWG tcon. David States is interested.

Karen: how can workgroups co-operate? The minutes from the last SDIWG ended in discussion of potential working groups http://www.na-mic.org/Wiki/index.php/SDIWG:Meeting_Minutes_20051021 Brian and David put four broad categories on the wiki. Four categories are: Software engineering; data integration and sharing; collaboration opportunities and challenges; infrastructure. Suzi says under ‘infrastructure’ that they will be putting up different services. These won’t be a basis for quadrification. Suzi wonders if we should split vertically across DBPs?

Aris: Summarize DBP: reverse eng networks B Cells; development of NLP framework; dissection of genetic determinants of autism and Alzheimer.

Chris Bliton UMich: Bipolar disease, diabetes I II, prostate C.

Peter Notes that Alzheimer is common. How to get a list of DBPs across all centers???

Aris: Software engineering is common effort. caBIG is good guide. Brian infrastructure. Leverage infrastructure within caBIG.

Brian: UMich comprehensive cancer center connected with caBIG.

Aris: Some of the items on wiki are dealt with in caBIG.

Dan Rubin: involved in imaging and vocabulary workspace.

Chris: Cell or anatomical imaging? Dan it has broad focus, but a lot is in clinical imaging. Chris: Robert Murphy CMU is cell imaging.

Barry Smith: shared notes for creating greater interoperability—so a general workshop may not have worked out, so decided to focus … ontology image ontology. March 24, 25: three goals: networking; ontology of images features etc; explore ontologies of imaging tools and data. Ivo Dinov … expand yellow pages … genuine classification. Advantages—can learn more about tools and how they play with other tools and data. Can bring together an important pair of groups imaging/ontology. Ontology of simulating tools as relates to underlying biology.

Andrea: how about expand focus to include interfaces for bioinformatics interoperability? Barry: Tools interoperability brainstorming. Andrea: Grid indexing ‘service’. Yellow pages… Brian: Covitz/Buetow/caBIG. Focused around “communities of use.” Mentions Suzi suggested that we think/organize around problem state of DBPs—do we have a list?? Andrea: Focus on SDIWG on IT part, but is it missing impact on biological and biomedical community—software has to be domain relevant. Brian: connects with Evaluation—how to measure Impact. Don Jenkins: clinical practice/treatment what impact can we make on that arena? Working with translational? Suzi says ‘How do we measure that’? Is compendium of software useful? Have to maintain it—overhead. Use case for ontology to see how ‘agents’ can locate. Barry: can feed into meeting.

Brian: at coming up meeting let each other know what offered up for further brainstorming. Don’t just do dog and pony show. Andrea: industrial side, one difference is that requirements are developed differently. May be waste of time to just put software out there…need to interrogate how bio-comput labs put software out there that works. Chris Bliton UofM: Mike Ackerman anthropologist and CS. Human computer interface. Collaborative environment. Community participation? Ontologies for DBPs. Suzi says must be community.

Peter: notes that BIRN and caBIG are represented here.

Karen: Move to agenda for Dec mtg.

Peter Good: Dog and pony is important—there will be NIH community there, so this is an important activity in consciousness raising. Other purpose is to begin interaction (Evaluation/SDIWG).

Karen: Morning can we grasp how the three new Centers (examples?) would build toward the networked centers—Suzi how to do it between now and then.

Peter Good: Centers are preparing their 5 pagers for the R01s.

Andea: metric for interoperability and …

Aris: reporting on ongoing activities—what are the pairwise events that have happened. What works!

Don: Center teams meeting without NIH staff is important. The strawman is for 1.5 hour PIs only. Thoughts from PIs? Flights from West coast get in 4 – 5 PM.

Aris: Open up floor after presentations. Three groups 1 hour a piece 8 to 12 (four hours). Can we shorten presentations? 45 minutes. Brian likes this. Maybe each Center just uses 45 minutes for their time – not break up Cores 1-3 and 4-7. Leave time for more detailed interactions. Organic interactions promoted around areas of mutual interest.

Karen: evening before reserve room for interactions.

Proposal: In the morning each group does 45 minutes dog and pony and 20 minutes open discussion.

Peter L: how to change the SDIWG direction of collaborative software environment? Brian: software eng industrial/academic. Focused working groups (see the list of four groups—we will work from that) http://www.na-mic.org/Wiki/index.php/SDIWG:Meeting_Minutes_20051021