ABI Meeting Minutes 2008-07-30

Present: Tommy Yu, James Lawson, Justin Marsh, Andrew Miller, Catherine Lloyd, Poul Nielsen, Dougal Cowan, Mike Cooling, Randall Britten, David Cumin

Apologies: Peter Hunter

Last week's action items

• No tracker item exists; an item should be created.
  

• This in progress, deadline for abstract submission is today
• Alan Garny is back on the author list for the tools / software submission

• The meeting was held before this one, and was productive. Mike will type up the notes he made for discussion at next week's meeting.

• James has gotten in touch with Penny
• Catherine mentioned that it is possible that a lot of people from the northern hemisphere who we would like to try the PMR2 prototype might be on summer vacation

Agenda items for this week

Plans for Singapore ICBME 2008 meeting

• Today is the deadline for abstract submission
  
• Andre informed us that a full length paper in addition to a poster / talk will be required for each accepted abstract. This may not be compulsory - more investigation is required
  
• The team is concerned about the impact factor of publishing our work in the conference's proceedings, rather than a full journal
• We also may not have enough material to write papers for some of the abstracts
• Poul was keen for people to submit abstracts anyway and if they are accepted, we can decline if necessary
  
• James, Mike and Sarala have submitted abstracts so far; the tools abstract and Tommy's PMR2 abstract will be submitted later today
• Sarala is almost certainly planning to attend the conference, and Andre will definitely be there
  
• We have put Andre and Sarala on all relevant abstracts so they are able to present the submissions if they are accepted
• Alan will be travelling through Singapore on his way home from New Zealand around the time of the conference. He has said that he would be happy to attend but that he would need help with funding for the conference fee and accommodation. This would probably be about $2000

Conferences that CellML team members might want to attend

• Catherine has put up a Google Calendar with conferences that we should be considering attending
• Catherine will be attending ICSB2008 and the SBML Forum meeting in Sweden in August
• Mike will be attending BMES2008 in St Louis, USA and SB4.0 [synthetic biology] in Hong Kong this October and MedSci2008 in Queenstown in November
• Poul noted that the Pacific Symposium on Biocomputing 2009 meeting in Hawaii might be worth sending someone to
• James attended an excellent Open Science workshop at BioSysBio2008. The conveners are to hold their next workshop at PSB2009.
• James is interested in attending the Biomodels training camp. [P.S. we were mistaken, the 2008 Biomodels training camp has already been held. We do not yet know if there will be a 2009 camp.]
  
• If anyone comes across any interesting conferences, please add them to the calendar

Little b

• Little b: "The little b project is an effort to provide an open-source language which allows scientists to build mathematical models of complex systems. The initial focus is systems biology. The goal is to stimulate widespread sharing and reuse of models."
• A paper describing little b was posted to CellML-discussion last week
• The group discussed their opinions on little b and whether they thought it might be able to achieve its stated goals
• The general consensus was that XML has significantly more support than LISP, and can be used to generate procedural code to do the majority of what a procedural language is able to
• James noted that they are essentially using a different approach to address the same problems with model sharing and reuse via modular componentry as we are; we should engage with them and seek their thoughts on this matter
• Mike also said that some of the problems with XML based descriptions that they are trying to address do not exist if a modeller follows the practices he described in his work with Peter Hunter and Edmund Crampin.
• Justin and Mike noted that mixing the solver and model as in little b is not desirable for a range of reasons (also mentioned by Herbert Sauro and Nicholas Le Novere in cellml-discussion.)
  

PCEnv progress update

• Both Andrew and Justin have mainly been working on bug fixes

CellML API progress update

• Justin has implemented a preliminary system that allows integrator points to fall exactly on defined boundaries. His commits are now in the trunk of the API, but not yet in PCEnv. See Tracker item 31.
  
• Andrew has almost completed the CellML validation service and is just fine-tuning it to remove false positives and errors
• He is considering creating a test-suite of models with specific errors to test his system
  
• He has run the validation service through the entire repository and has produced a large list of what models contain errors and what they are
• This service will eventually go into PCEnv. In the mean time, it may be worth making a command-line tool available

CellML 1.2 specification

• Andrew has had no feedback on his proposal for the lambda-calculus based typing system for CellML 1.2
• Poul would like to read it but has not had time lately. He will be away next week, but will read it the week after
• Mike also had a look but has not had time to finish reading it
  

CellML metadata specifications: core and graphing & simulation

• The core CellML metadata specification has been referenced a lot, and has been used as the basis for SBML metadata, but is not actually a finalised specification
• Andre suggested we finalise the metadata spec during his last visit to Auckland
• James is also very keen to see it finalised. Tommy noted that it actually needs a complete rewrite
• Tommy suggested that we should let the CellML metadata spec focus more on CellML elements, and let citations be handled by leveraging other systems
• There are a number of new standards and technologies that we will be able to leverage. BQS (bibliographic query system,) for example, which is not in wide use and has a number of issues, can be replaced with the MIRIAM bibliographic citation system
• Randall does not believe that we currently have the resources to devote to working full-steam on rewriting the metadata specification. Sarala might be able to put some time into this after she finishes her PhD
• James is keen to start getting momentum up, even if interested parties just start doing a little research and making notes when they have time spare. If we are to start putting significant energy into annotating models, we really need a static metadata spec
• Randall suggested that the Graphing & Simulation Metadata specifications are a priority
• Justin is very keen to start devoting some of his time to this specification
• Poul asked what the status of graphing and simulation metadata is within the SBML community
• We don't really have a good idea of this, but Andrew mentioned MIASE, which is apparently supposed to be format agnostic, but appears to be quite specific to SBML so far
  
• When they distributed their specification initially, Andrew was concerned about their assumption that time be the only x-axis value possible. They did not make any response to his comments. MIASE also lumps graphing and simulation together, which we do not believe to be appropriate for CellML
• We should keep an eye on MIASE and see how it develops. Poul doesn't think it is worth making our graphing and simulation metadata MIASE-compliant unless the standard has a significant following
• Andrew noted that we should ensure that we are implementing the draft Graphing & Simulation Metadata specifications as they are now

PMR2 progress update

• Tommy has been working on presenting graphs generated from the CSV output of models within the model page in PMR2
• He has also been working on the infrastructure for the curation flag system

Update on additions of content to repository

• Catherine has added a model for Paul Harrington
• She will be meeting with Kumar Mithraratne and Paul to discuss using CellML
• Paul has several models suitable for CellML, but they will need to be imported into OpenCMISS to be run.