ABI CellML Meeting Minutes, 12th January 2011

Present: Andrew Miller, Dougal Cowan, David "Andre" Nickerson, Tommy Yu, Amy You, Poul Nielsen, Peter Hunter

Last week's action items:

This week's agenda:

1) (Andre) CellML workshop

Discussion:

Andre: the workshop will probably be held on the week beginning on the 11th of April, on the Monday and Tuesday. Alan has been emailed and is happy with the dates. The room now needs to be booked, which shouldn't be a problem as this is the first week of mid-semester break.
Andre: Focus on current issues on one day, and then future requirements on the second day.

Planning:

Poul: need to get some requirements proposals before the meeting.
Andre: should be presentations on proposals, leading into discussion. This will help us get an idea of what to decide for 1.2, and set a timeline.

2) (Alan) For OpenCOR: PMR2 services and CellML API binaries

Discussion:

Alan: will PMR2 provide web services through which OpenCell can clone models, search, etc?
Tommy: yes, see tracker item 1909.
Alan: will binary API builds become available?
Andrew: "informal" binaries are currently available, and with the next API release these binaries will be available as standard.

3) (Andrew) CellML API survey

Discussion:

Dougal: there have been 6 responses so far to the survey, with some useful information.

Planning:

Andrew: should we send another email, and set a deadline?
Andre: should be send this to other mailing lists, such as SBML?
Lukas: there are people on the SBML mailing list who could well be interested.

Action item 1: Dougal to send an email asking Mike Hucka if it is appropriate to send the survey email to the SBML mailing list. Then send a reminder email to the appropriate list(s), with deadline of max 2 weeks.

4) (Andrew) DVCS paper

Discussion:

Andrew: The paper has been accepted in principle, but needs some very minor editorial changes. Everyone has to be happy with the final version with these changes before it is sent off.
There was a discussion about the issue with referencing additional files.

5) (Andrew) Stochastic parameter variation

Discussion:

Andrew: how do we want to implement stochastic parameter variation? Should this be part of the model or part of the simulation metadata?
Peter: the model should contain the description of the PDF or cumulative probability density (possibly by importing an appropriate CellML library module) for parameters which have uncertainty, but the simulation (SED-ML) metadata should describe the process of selecting an instance of the parameter value from the distribution using a random number generator.
There was a discussion about the two classes of stochastic models - 1, where the parameter has a true value but this value is uncertain; 2, where the parameter varies stochastically.
Poul: need a way of saying that a parameter has uncertainty, and then SED-ML has to recognise such variables and apply a distribution. We will need a library of CellML models for cumulative probability functions.
There was a discussion about this library.
Lukas: looking at the SBML proposals for stochastic modelling (Darren Wilkinson's), these could be very helpful.

6) Repository contributions - update

Progress:

Catherine has put the following models into the repository, which all await diagrams:

Fujita, Toyoshima, Uda, Ozaki, Kubota, Kuroda, 2010
Decoupling of receptor and downstream signals in the Akt pathway by its low-pass filter characteristics

Morris, Lecar, 1981
Voltage oscillations in the barnacle giant muscle fiber

Wang, Cirit, Haugh, 2009
PI3K-dependent cross-talk interactions converge with Ras as quantifiable inputs integrated by Erk

Alexander, Wahl, 2010
Self-tolerance and Autoimmunity in a Regulatory T Cell Model

Smith, Abdala, Koizumi, Rybak, Paton, 2007
Spatial and functional architecture of the mammalian brain stem respiratory network: a hierarchy of three oscillatory mechanisms

Shrestha, Hollot, Chipkin, Schmitt, Chait, 2010
A mathematical model of parathyroid hormone response to acute changes in plasma ionized calcium concentration in humans

Mona has put the following models into the repository:

Leloup, Goldbeter, 1998
A Model for Circadian Rhythms in Drosophila Incorporating the
Formation of a Complex between the PER and TIM proteins

Leloup, Goldbeter, 2004
Modeling the mammalian circadian clock: sensitivity analysis and
multiplicity of oscillatory mechanisms

Leloup, Gonze, Goldbeter, 1999
Limit cycle models for circadian rhythms based on transcriptional
regulation in Drosophila and Neurospora

7) PMR2 development - update

Progress:

Tommy: the PMR2 paper is in advanced publication.
Tommy: short URLs are now the default setting. Did some further refactoring and work on tests.
Andre: has the embedded workspace listing been implemented? Tommy: no. Bigger features will wait until the 0.4 release.
Andre: a note needs to be added so that people who use the repository know to cite the paper.

8) CellML API - update

Progress:

Andrew: TeLICeMS tested and debugged, now performs round-tripping.
Andre: can TeLICeMS just process a part of a model? There was a discussion about this.
Andrew: ported Lucian's code to TeLICeMS, so it no longer uses XPCOM. Helped Lucian get this code up & running. There are now no XPCOM API users. Fixed the TeLICeMS build system and code generator, bugs found by Lucian. Some memory bugs in TeLICeMS also fixed.
Andrew: tracker items 2789 (mulitpliers), 2731 (test suite for 64bit), and 2759 (custom code generation) fixed. A TeLICeMS bug with piecewise equations without otherwise fixed.
Andrew: looking at how we would implement SED-ML in the API.

Discussion:

Andre: we should be using existing SED-ML tools if possible.
There was a discussion about using the SED-ML API(s), existing tools, etc.
Peter: get in touch with other SED-ML developers and see how they can help and how Andrew can contribute to their efforts. It is important that we don't go off on a tangent.
Andre: we should be raising issues with SED-ML on their mailing list and keeping in touch with them rather than developing our own tools or API.
Andre: need a runcellml excecutable on Windows - is it available? Andrew: no.
Amy: Frank is waiting on a command line CellML simulator for Windows and Mac.
There was a discussion about providing a CellML exe of some kind on Windows.
Peter: David Bullivant says that the hardware is ready for the Windows build system.

Action item 2: Andrew: produce a runcellml.exe for Windows.

Andre: Alan had put a GPL license on the new OpenCell, but is going to change this and wants input on the choice of license. Currently favouring the BSD 3 clause license.

9) CellML core specs - update

Progress:

Andrew: ModML - the Rice et al. 2008 model is now working, reproducing the results from the XPP version. Work is underway on the Tran et al. 2009 model. There are issues with different versions of this model in CellML having mistakes or missing equations.
Andrew: the Tran et al. 2009 model requires subtractive modelling. To support this, ModML will need to filter out some processes and replace them.
Andre: may have a correct working CellML version of the Tran et al. 2009 model.

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

Poul: if people want final year Software Engineering students, projects need to be sorted out fast.