ABI Meeting minutes 2009-02-11

Present: Randall Britten, Dougal Cowan, Adam Reeve, Andrew Miller, Catherine Lloyd, Poul Nielsen, Justin Marsh, James Lawson, Jeelean Lim, Jonna Terkildsen, Tommy Yu, Peter Hunter, Mike Cooling

Last week's action items:

1.) Catherine and James to draft a model request submission form.
  • Catherine has created a news item asking for model request submissions.
  • James has created a form.

This week's agenda:

1) PCEnv/COR name. Update.
  • There was a discussion of the voting results between Andrew and Peter; it was noted that there were no votes from Oxford.
  • Peter Hunter wanted a list of the "top 5" names as voted for so far, so that there can be another vote on this reduced list of names.
  • Andrew did not see the purpose of a reduced list, due to the voting system employed.
  • Catherine and Randall suggested that people may be tired of the whole process, and that a large list of names can be fatiguing to vote on.
  • Mike suggested that we could use this vote as a dry run to decide the final process.
  • Peter Hunter suggested the name "CYTOSS", and said that Alan doesn't mind this name.
  • Catherine said that if all of the developers are content with the name CYTOSS that we could perhaps just go with that as a final name.
  • Andrew asked what hits CYTOS has - as it is a biomedical company, Andrew thinks this is a strong objection to the name.  It was also suggested by Poul that it would be hard to spell CYTOSS with one s.
  • Poul thinks that OpenCell is fine, but that the domain ownership issue is a big problem.
  • Peter sees a reasonably clear majority in favour of OpenCell.
  • James suggested (in light of the low voter turnout) that we might need to look at the way we engage the community with these decisions to improve participation.
  • Poul said that we need to know the reasons for people not participating, rather than just presenting a process and saying that "they've had their chance" if they do not take part.  Poul also said that there are people who we definitely need to have involved, and if they don't get involved then the process has failed.
  • Randall suggested that we move on and return to this item later.
  • Peter wants a few more days before a decision is made.

2) SBML to CellML converter  - update.
  • Catherine suggested that Andrew call Nicolas Rodriguez.
  • Andrew reported that the model provided by Catherine exposed some bugs in the converter, and is also very slow (minutes).
  • Catherine's model is a very new glycolysis model, currently only available in SBML, from the Manchester group.
  • Andrew has been working on bug fixes and optimization for speed.
  • Randall explained that slowness is a major problem for development as it makes testing very laborious.
  • Andrew is almost ready to give Catherine a CellML model from the SBML model she provided.

3) Website redesign - update
  • Catherine reported that there was no meeting this week.
  • James has been migrating content and re-writing old content.

4) PMR2 update
  • Tommy has been finishing the text transformation to exposures, and working on exposure deprecation by new exposures.  There will be explicit relationships between old and new exposures.
  • James asked if it is going to be possible to branch exposures; Randall explained that you will be able to make multiple exposures and explore the exposure history.
  • Tommy mentioned tracker item 1589.
  • James said that we need to develop use-cases.
  • Randall said that people should look at the tracker item - he expects a lot of discussion on this tracker item, especially from Andre.
  • There was an extensive discussion on the versioning and exposure capabilities of PMR2.
  • James brought up the idea of some kind of page that documents collections of models and variants.
  • Randall stated the importance of keeping the scope of PMR2 such that Tommy can finish by the workshop - he suggested a breakaway meeting to discuss PMR2 features.

5) PCEnv development - update
  • Justin reported that Litmus testing is currently happening.
  • Catherine has found bugs in the Mac version.  She reported a bug that prevents scrolling in an expanded tree view.
  • Peter Hunter asked what the new features are for 0.6.
  • Justin mentioned Mac support and the new preferences functions.
  • Randall mentioned the MATLAB and Python code export functions.
  • Peter asked if anyone is using Protege editor to edit annotations, and asked people to think more about metadata editing.
  • James said that you need Sarala's tool to export OWL for editing in Protege - he asked if Sarala has done any more work on this. Poul said that she probably hasn't, as she has been writing papers.
  • Peter asked to have a meeting within the next week after people have revised their knowledge of metadata editing.

6) CellML API - update
  • Adam is working on the CeLEDS export service.
  • Andrew has been working on bug fixes - pretty much ready for release now.
  • Randall brought up the issue of the mis-named function.
  • Andrew is reluctant to change it to avoid breaking things.
  • It was generally agreed that now is the best time to change it if it is going to be changed; Randall suggested having both the mis-named and re-named functions in first, then deprecating the old one.
  • Randall suggested a breakaway on the renaming of the function.
  • Poul mentioned students coming from UCSD to check out the Java API so that they can update the continuity support of CellML.
  • Mike mentioned that the testing of the Java wrappers is very slow; he is currently setting up the build environment and has limited time to work on it.
  • Andrew asked if Mike is planning to document the process; Mike says definitely!
  • Jonna brought up the issue of one of her models failing on first attempt after 10 to 15 minutes of taking random guesses (after the 99 random guesses fix).  She wants to be able to set initial conditions, and would also like a status message along the lines of "Taking random guesses for initial values" so that the user knows what is going on.
  • Randall suggested that we need to provide more detailed feedback on the integration process.
  • Andrew asked if we are going to branch soon; Justin wants to wait for the Litmus results.
  • Jonna asked how the additional random guesses are being scaled; Andrew said by number of variables.
  • Randall proposed a breakaway on choosing the scaling factor.

7) CellML specs - update
  • James is starting to write up the specs in the form of the current (old) specs, in a largely informative style.
  • James asked if he has to use git to version the specs?
  • Randall stressed the importance of a version history, whatever system is used.  James said he will probably use git in this case.
  • James asked about the updated graphing and simulation metadata.
  • Randall said that the scope for PCEnv 0.7 has not yet been decided.
8)  Repository contributions - update
  • Catherine has fixed the Pasek 2006 model.
  • Pulasthi has been doing diagrams, Linda has been making sessions.
  • Jeelean has been working on cell cycle sessions.
  • James is going to give another session on session making.

9) Updates or additional items
  • Catherine suggested that before the workshop, everybody get some background knowledge on BIOPAX, MIASE, SBML, SBGN etc.  She suggested a meeting at 11am on the 18th of February for this purpose.
  • Peter showed some VPH NoE newsletters, that contain an announcement about the workshop.