ABI CellML meeting minutes 2010-02-10
Present: Randall Britten, Catherine Lloyd, David "Andre" Nickerson, Poul Nielsen, Justin Marsh, Mark Hanna, Ethan Cho, Tommy Yu, Dougal Cowan, Tessa Paris, Andrew Miller, Mike Cooling
Apologies: Peter Hunter
Last week's action items:
1) Tommy to add a "how to cite" link to the PMR2 list.
- Not done.
2) Catherine to create a tracker item about connection creation.
- Done. Alan and Justin have commented on the tracker item (2409).
This week's agenda:
1) (Dagmar) Model URL inconsistency is causing a problem when trying to cite particular models in a paper manuscript
- Catherine reported that this URL inconsistency has been fixed. There is a call from users for a simpler URL scheme, as requested by Peter.
2) (Dougal) What information do we want in the CellML tutorial at the workshop?
- Catherine said that we need to discuss and decide what to do for the workshop CellML tutorial. This could be a very simple OpenCell tutorial, and then a PMR2 tutorial.
- Andrew suggested checking the list of attendees, as maybe this tutorial will not be relevant to them.
- There was a discussion about who might attend the tutorial.
- Andre suggested using the same tutorial as was used at IUPS in Kyoto.
- Mike said that he is happy to do his simple tutorial.
- There was a discussion about the workshop, time spent on tutorials, etc.
- Andrew asked if we have any idea about working groups for the hackathon.
- Mike asked if it is worth putting a slot in to cover issues discussed at meetings with Sarala - Poul said he could cover these in his talk.
3) CellML core specs - update
- Andrew continues to work on the converter, from base ModML into IDA code. It is nearly at a stage where models could be written in base ModML.
- Andrew said he will need to talk to Mike about his work.
- Randall asked if ModML will require a new API. Andrew said that the demo will have its own API, but that does not mean we're stuck with this.
- Mike asked if there could be a demo for the workshop. Andrew said that it will only really be possible to cover the concepts at the workshop.
- There was a discussion about finite delays, and the possibility of getting a student to work on this as a project.
4) Metadata specifications (general, citations, graphing, simulation, other)
- Randall again mentioned the need to look at SEDML for OpenCell 0.8.
5) cellml.org update
- Only the workshop pages have been updated.
6) OpenCell development update
- Justin has been working on the keyboard tree navigation, which is largely fixed with a new version of xulrunner - however this introduces other problems.
- Justin reported that the CSV export now gives full paths for variables.
- Andrew has been working on the Windows build environment installer. He has started work on this using NSIS instead of Ruby Shoes.
- A breakaway was called to discuss the Windows build environment installer.
7) CellML API - update
- Andrew reported that Alan has made some minor changes, which now need to be approved by the other authors.
- There was a discussion about changes to the paper.
8) PMR2 development - update
- Tommy has the cmgui demo done for Peter, which should be hosted tomorrow.
- Tommy continues work on github (user-prefixed) style URLs.
- Tommy has been refactoring PMR2 code.
9) Repository contributions update
- Catherine has been working on endocrine models, going through the whole list.
- Tessa is working on immunology models; she has made a T cell and a couple of cancer models.
- Ethan has finished off a Rong model, and was working on another one that required a distribution and can't be done in CellML.
- Randall asked what the distribution was doing in this model - it was randomly setting a variable to 1 according to a normal distribution.
- There was a discussion about use of distributions in models.
- Poul said that this has cropped up before, about whether CellMl can handle variables that are stochastic.
- There was a discussion about including probability distributions in CellML, where to include these, as inputs or functions, etc.
- Randall asked if papers the CellML repository models are based on specify uncertainty for variable values, and whether it could be useful to capture these uncertainties with distributions.
- Poul said that it might be useful to have structured types - it would be nice to be able to specify a variable that is tied to a distribution, for example.
- Mark, Hanne and Dougal have created a number of SVG sessions for electrophysiology models.