ABI CellML Meeting Minutes, 21st July 2010

Present: Dougal Cowan, Randall Britten, Andrew Miller, David "Andre" Nickerson, Tommy Yu, Justin Marsh

Apologies: Peter Hunter

Last week's action items:

1) Randall - ask Mike Cooling about containment in metadata.

  • Randall has been in touch with Mike about this, and Mike is looking at it.

2) Justin - find out what version of the API and model Geoff was using by looking at his machine/VM.

  • Not done, but will be done today.

This week's agenda:

1) Repository contributions - update


  • Penny has updated several cardiac electrophysiology models which are now running at steady-state.  Further changes were made to some of the models and all of these alterations have been documented in the PMR2 commit messages.
  • Hanne: continuing to create SVG diagrams and OpenCell session files for the curated neurobiology models.
  • Catherine: encoded a multi-cellular neuronal model (Butera et al. II 1999) using CellML 1.1.  She's now systematically going through the metabolism section curating the models which haven't been looked at since PMR0.

2) PMR2 development - update


  • Tommy: working on the curation module. Has been cleaning up the codebase for interoperability with Plone, some of these changes are required for the curation module.
  • Justin: has done a lot of work on tracker item 1703, the e-notation fixing script.


  • Andre: The current e-notation fixing script changes a lot of the text formatting around namespaces.
  • Andre: Hugh Sorby has helped out with a sed regex script which only changes <cn> elements and makes no changes to the XML formatting.
  • There was a discussion about sed vs tool-based fix scripts.
  • Andrew: has had feedback from Alan about the paper, and these changes are now processed.
  • Andrew: needs to know if anyone else has any changes to suggest.

Action item 1: Tommy - edit the authorship information in the application note using MS Word.


  • The work done on the e-notation script will be put on hold for the time being, although this code may be useful in future.
  • Randall: Suggested that we shelve the idea of doing batch conversion, and make both the regex script and Justin's Scala/Java tool available for curators to use on a case by case basis.

3) Metadata specifications (general, citations, graphing, simulation, other) - update

  • No update.

4) CellML core specs - update


  • Andrew: received an e-mail from Poul work done by Robert Muetzelfeldt on "Handling structural complexity in System Dynamics models" and the Simile environment. Andrew is reading the paper and will reply to Poul.


  • Randall: could ModML be done in Scala?
  • Andrew: ModML could be done in any language that supports functional programming, but the fact that Haskell is a pure functional language helps to ensure models created are declarative in nature.

5) CellML API - update


  • Andrew: finished a feature to allow "custom code generation", where the user can select computation targets that are known, wanted, or unwanted.
  • Andrew: has notified the OpenCMISS developers, and also the Continuity developers (via Andrew McCulloch) of this feature, and asked for feedback on the API that has been developed for it.


  • Randall: could custom code generation be announced more widely?
  • Andrew: this feature could be announced. An application note could be written on this feature.
  • Andre: the API needs user documentation.
  • There was a discussion about user-documentation of the API.
  • Andrew: Reference documentation of the API already exists.


  • Randall: This would meet the requirement: "user guide" style documentation to the API, i.e. documentation for possible usages of the API, e.g. how to use the API to generate code; how to use the API to support the development of a GUI modelling tool. Documentation for usage from C++. Separate documentation showing how usage from C++ maps to usage from other supported languages.

6) OpenCell development


  • Justin: working on the non-admin install for OpenCell on Windows.


  • Justin: is it worth spending a day on the non-admin install?
  • Randall: yes.
  • Randall: IDA support is now the priority - getting some repository models to work with IDA is important for 0.8.
  • Andre: a simple model that doesn't work with IDA should be the target to fix, rather than something complex such as Saucerman.
  • Andrew / Justin: Will try to get in touch with Peter while he is in Auckland to discuss the use of the Scala platform for OpenCell.