ABI CellML Meeting Minutes, 16th of March, 2011

Present: Dougal Cowan, Randall Britten, David "Andre" Nickerson, Andrew Miller, Mike Cooling, Tommy Yu, Lukas Endler

Apologies: Poul Nielsen

Last week's action items:

[From 2 weeks ago]

1) Randall - will put Mike and Sarala in touch on this issue. [RICORDO]

  • Done.

2) Tommy will check that there is an RSS feed [of models to be published], and add one if necessary, and then Randall will check in with Dougal.

  • Done.

This week's agenda:

1) SI/CGS unit compatibility in JSim (Andre)


  • Andre: Eric Butterworth has enquired about what possible use of CGS units (ESU) in Jsim could mean for interconversion to/from CellML. We could define these units in CellML, but they are not compatible with the CellML SI units.
  • There was a discussion about CGS and SI compatibility, possible conversion techniques, and unit definitions.
  • Andrew: the only level of support we would want to offer would probably be a table of conversions. As Jsim has user-defined units, could they not simply define a library of CGS units?

Action item 1: Andre to contact Eric to explain our thoughts about this.

2) CellML API survey results - discussion


  • Dougal: gave a brief summary of the results, with some comments and the results of the feature priorities.

Action item 2: Andrew to summarise the results of the survey for next week's meeting.

3) Metadata draft spec - discussion


  • Mike: Sarala and Pierre are developing a system where the owners of models submit their metadata, where it is stored prior to querying. It is up to owners of models to push the appropriate update, delete, create information incentivised by the prospect of being able to execute said queries. All elements and strategies are someone prototypical at this stage, and are subject to change.
  • Tommy: will RICORDO allow both updating and removal of metadata?
  • There was a discussion about updating and removing metadata.
  • Randall: perhaps the metadata could be updated on exposure, and removed at expiry?
  • There was a discussion about how to reference model files in various locations, on various devices and at various URIs, etc.
  • Andre: should we register a MIRIAM resource for exposures, to provide a MIRIAM URN?
  • Lukas: could this be done like the reactome, which has identifiers for versions and access to files within versions? (?)
  • There was a discussion about SED-ML, identifiers, etc.
  • Mike: doesn't like the addition of cmeta:ids during annotation, as it changes the model file. Could xpath be used to point to locations within the model?
  • There was a discussion about xpath, cmeta:ids, and the standards compliance issues of using xpath.
  • There was an extensive discussion about various ways to annotate models.

4) PMR2 development - update


  • Tommy: started integrating Andre's CellML Simulator into the back-end. The front-end will need more work. There is a security hole which needs to be fixed, allowing the execution of arbitrary JavaScript.
  • Tommy: finished refactoring, breaking up the core packages.


  • There was a discussion about HTML uploads into workspaces, the security issues, and how to fix them.

5) Repository contributions - update

  • No update.

6) CellML core specs - update

  • No update.

7) CellML API - update


  • Andrew: working on unit tests for SProS - there is one test left to write.


  • Andrew: need to make something now that uses SED-ML  (?).


  • Randall: Chris Bradley has reported that a milestone has been reached in using CellML in OpenCMISS.

8) Time permitting: Updates or additional items from volunteering attendees

  • Randall: the CellML summer tutorial went OK.
  • Mike: the CellML team on the website is out of date

Action item 3: Dougal - update the team page on cellml.org.

  • Randall: would like to see dates on the website pages. Tommy will change the template to display the last modified date for all items.