Meeting Minutes 2008-01-09

Present: Randall Britten, Michael Cooling, James Lawson, Catherine Lloyd, Peter Hunter, Justin Marsh, Andrew Miller, Tommy Yu

Apologies: Poul Nielsen


Agenda

Alan: CellML workshop/meeting 2008 update

  • Catherine has a list of confirmed attendees and the list of twelve speakers and the topics they will cover.

  • Peter named a few other people that may want to speak on a few different topics.

  • Mike noted that he can speak about using CellML as he has had some experiences that he might be able to share.

James: Awareness of CellML in ABI

  • James pointed out that the CellML group has grown significantly during 2007 and is now one of the larger groups within the ABI. However awareness within the ABI of work relating to the development of CellML and related tools needs to be improved.

    • Peter pointed out that one of the ideas raised during the ABI retreat was for people to be able to give talks. The proposals raised during the annual ABI retreat was to split the regular Friday morning seminar session into three 20 minute blocks. James suggested that the CellML team should make use of these shorter blocks and present regular talks during 2008.

      • Michael would be keen to volunteer to give a Friday morning talk with regards to thesis work using CellML.

    • Curation of models created here will be of help.

    • Peter noted that fostering closer relations between the CellML team and various modelling groups within the ABI could help to improve awareness of the potential relevance of CellML to their research. The possibility of engaging one member of each group and helping them build their knowledge of CellML was raised; these people would then be able to teach their colleagues about CellML.

  • To foster a good environment to exchange models, the model repository needs to support a more robust access control.

    • This will include reviewer access. It was noted that the current PMR handles this insufficiently due to the way the permissions settings were set up on the site.

    • Currently Tommy is taking the time to learn Plone 3 and Mercurial, which are not simple technologies.

    • Randall pointed out the design goals that were discussed with Tommy yesterday that the PMR2 would need to cover.

      • This includes support for CellML 1.1, workspaces, curation and replication

      • The target is to get some sort of prototype working that covers the major planned features.

      • It was decided that functionality (especially support for CellML 1.1) was more important than the access controls for the prototype.

    • However it was noted that we do need to set some sort of timeline to complete this prototype. The workshop was decided as a natural time and it could be possible that Tommy gives a talk on what the next version of PMR will do and to exhibit the prototype.

      • Andrew noted that with all these features, it might be best not to enforce a too concrete timeline.

      • Catherine and James volunteered to discuss about curating CellML models if Tommy cannot give the talk in that time, however the actual schedule for the talks is quite tight already.

Randall: PCEnv 0.4/CellML API 1.4 priorities

  • There was a discussion with PCEnv/CellML Api developers on Tuesday.

    • Mac release for PCEnv was scheduled as top priority (when planning the PCEnv 0.3 release, Mac support was deferred, under the understanding that it would be worked on straight after the release so that PCEnv 0.4 was to have a working Mac version), but discussions with Peter Hunter later on Tuesday and via developers mailing list resulted in that being replaced with the task of adding support for solving multidimensional non-linear algebraic systems during integration.

    • It was also noted that we are not certain what the demand is for PCEnv on Mac.

      • Andrew pointed out that it the importance depends not so much on how many current Mac/PCEnv users there are, but rather how many potential users.

  • The outcome was the following plan
    • Justin's next tasks:
      1) Non-linear algebraic system solver 
      2) Continue Andrew's work on CellML validation service
    • Andrew's next task: Continue work on CellML specification version 1.2
    • Alan's next task: Refactoring of the user interface, but it probably won't be ready for 0.4 release.
  • Peter gave an explanation of what a non-linear system is, what PCEnv can currently do (just solve a single non-linear equation but not more than one non-linear equation).
  • A possible implementation approaches were discussed:
    • User could be required to flag non-linear algebraic system (NLAS) via metadata.  
    • Software should either notify user which equations are circular, and may need to be manually flagged as part of the NLAS.
    • NLAS equations could be automatically detected by the software (Andrew's pointed out that Poul favoured this approach, since it is cleaner.  Andrew also pointed out that the flagging approach creates the burden of another metadata specification.)

Peter on the European Framework

  • The ABI has joined the network of excellence in the European Framework, and is currently the only non-European participant.

    • CellML and FieldML are going to be used extensively by all European groups for modeling

  • There are areas that have been granted funding

    • Heart modeling.

    • Drug design

      • Designing a better measure for effectiveness in the heart than the current practices.

  • James – with new people and projects, the website could use an update

    • Peter – there will be a full time staff that will focus exclusively on documentation.


Updates from Everyone

Catherine

  • Looked through the model repository over the break and recoded eight models that used reaction elements.

  • There are still eleven outstanding models that had the deprecated reaction elements.

    • Many of those models are "learning" models that were built using pathway diagrams.

    • Question was raised on whether it is fine to have a model without mathematics.

    • Action item: Randall will add a note on next week's agenda to invite Poul to give his input on this matter.