ABI Meeting Minutes 2008-04-09

Present: Tommy Yu, James Lawson, Randall Britten, Andrew Miller, Catherine Lloyd, Justin Marsh

Apologies: Peter Hunter, Poul Nielsen

Review of last week's action items:

1.) James will email Matt and ask for the Protégé files.

  • James emailed Matt, who said that although Mikey Lovell-Smith's work may no longer be particularly relevant to the task at hand, he would find the files and email them to James when he has time.

2.) James will do some research into how we should approach this work and come up with a proposal suggesting the direction we should be taking.

  • James has started on this task and met with Sarala to discuss her current ontology framework and how to begin experimenting with ontological annotation.

  • It is possible to import SBO or to simply reference entries in it directly by their URL. Therefore most of the terms within a CellML constrained vocabulary should be terms that can not be found elsewhere. This work is ongoing.

3.) We will have a meeting to discuss the future of the graphing and simulation metadata specifications while David is still here.

  • Justin, Randall, James and David attended the meeting. Randall felt that the discussion was useful but we did not make any explicit decisions.

  • David suggested that while we are now working on CellML 1.2 specification, the draft metadata specification is yet to be officially released. Perhaps we should put more effort into finalising the metadata specification.

  • James noted that in his investigations into the differences between CellML and SBML for his upcoming talk at the University of Newcastle, it has become apparent that metadata in CellML is essential to their view of the language.

  • Catherine noted that she would be interested in James's investigations into CellML / SBML comparisons, as she is sure people will ask her about it after her conference talk in Amsterdam. Andrew has also been working on factoring this into his talk in Seattle at the synthetic biology standards workshop. He, Catherine and James should get together and discuss their talks.

Agenda for this week:

1.) Andrew - Review poster for synthetic biology workshop

  • Andrew has solicited input from members of the CellML community on his poster describing the tools and software available for working with CellML.

  • An issue raised by Randall and James was that of the model Andrew has chosen to show in his repository screenshot. He has chosen a pathway model, but one with only one curation star. It was suggested that he should display a model that has been better curated. James said he would let Andrew know what models might be better to use.

2.) Catherine - CellML website content - feedback and monthly meeting

  • Catherine met Mark Gahegan at the CellML Workshop. James and Randall also chatted with him. Mark is a Professor of Geography and has done a lot of work in Geographic Information Systems.

  • Mark was particularly interested in the cellml.org website and the CellML project for its strong emphasis on engaging and growing community, and would be happy to be involved in a consultative role in the future development of the site and web-based community.

  • Catherine had a meeting with Mark, cellml.org webmaster Gareth de Walters and University of Auckland e-research project manager Nick Jones to discuss the design vision for the website.

  • Gareth suggested that a design committee be created to work on cellml.org and plan its future.

  • Catherine suggested that the committe meet monthly and the team discussed who might be on this committee. Suggested members were: Tommy, James, Catherine and Gareth.

    • David Nickerson is also likely to want to be in the loop on these matters, but usually prefers a typed up summary to joining meetings by conference call from Singapore.

  • The team then discussed the mandate of this group and consultation via the cellml-discussion mailing list on decisions affecting the community was encouraged.

Action Item 1.) Catherine will send an email suggesting a time in early May after she and James return from overseas for the first meeting.

3.) Randall - Discuss "PMR2 parallel workspace use case": tracker item 371

  • There will be a breakaway meeting to discuss the technical aspects of this issue.

  • When a user creates a CellML 1.1 model, import URLs that refer to locations on the user's hard disk need to be changed to refer to the appropriate repository URLs when the file is uploaded to the repository.

  • If nothing is done about this situation, and the model is a tree-like composition of many branches, a great many URLs may need to be changed when the model is uploaded.

  • The user will also need to predict what the repository URLs will be, and will not be able to test the model from their hard drive after they have changed the URLs but before they upload it. If they are changing many URLs they will almost certainly make errors.

  • Therefore Tommy and Randall believe that there should be some software based mechanism to alleviate this problem for CellML 1.1 users, lest people find it too much work to upload models to the repository.

4.) PCEnv / CellML API

  • Randall has monkey-patched PCEnv to allow the latest snapshot Windows build to display MathML in normal text, rather than Greek.

  • Mozilla has fixed this bug in their code, but a creating a Linux snapshot of PCEnv that incorporates their latest build is proving difficult.

  • Justin has fixed the rendering of incorrect and extraneous lines in graphs in PCEnv. He has completely changed the way sampling for the rendering is done, the trade-off for which is a moderate slowdown in the response to graph panning.

  • Andrew will be taking over the project to implement the non-linear solver in the API.