On Tue, 17 Jul 2007, Maciej Swat wrote:
Dear Jim, James Glazier has asked me to send you a copy of Model Sharing Mini-Workshop summary so that you have it ready when working on modeling standards document. The document attached essentially summarizes recommendations made at the workshop as far as what is required to improve model sharing . I am sure that most of the material in the document you are working on is already in-sync with the recommendations made at the workshop, but it does not hurt to send you a copy of it. Also, if you find that some of the recommendations are questionable , please let us know so that we can discuss it with people in our working groups If you need anything else from us please let me know ...Maciek (JBB: The report is the previous entry on this Discussion Page.)
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
from Jim Bassingthwaighte" 19jul07:
Got the report on the workshop dicsusion of 10 April07. Thanks.
Good, however I really disagree with part of the report, namely the implicit view that SBML and CellML are acceptable standards, Of the 5000 SBML modelson the site, only a few are actually correcxt enought to run when translated into JSim. The Biomdoels dbase, in contrast, almost all of them run, and the three that don't may be JSim problems rather then their curation probelms.. we don't know yet and have to examiine the failures.
Likewise, a small fraction of CellML model run, even the ones supposedly curated, though we are workiing with them on some of these. My pont is NEITHER CellML or SBML are acceptable standards for dissemiination. They are simply repositories of untessted pieces of code, for the most part and do not set a standard. They should perhaps be designated as holding tanks of models awaiting curataion, and then be moved to a distinguished site when really curated.
Herb, Dan, do you agree with my rude statement? ...Jim
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
from James Glazier, 19jul07
Our write-up was simply a summary of what people said at the meeting--it does not represent "our view" or anything more than the coallated comments. A new approach that goes beyond SBML/CellML is definitely a great idea. ... JAG
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
from Dan Beard, 20jul07:
I bet you are right about the failure of CellML and SBML to meet the standards of Jsim. I wonder if it is worth defining different classes of standards. (Maybe the report that you are talking about does this, but I have not seen the report or know where it comes from.
The ML's define very useful (but still imperfect and evolving I guess) computer science-based standards for representing models. These are standards for how information is represented and what is the necessary information to define a mathematical model. The idea is to have a framework that is precise (no ambiguity in model specification), but maximally general (anything can be made to fit to the standard).
I suspect that Jsim is detecting (at least in part) scientific mistakes such as unit imbalance. This gets back to an old discussion that we have had with the CellML people. One thing that I took away from that discussion is that the ML developers are not going to worry about scientific standards at the level of the ML's. Maybe this makes sense if the applications tools (i.e., Jsim) take up the burden of introducing and maintaining scientific standards.
Make sense? Or is this too simple-minded of a way of thinking about the issue? ..... Dan
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
20 Jul 2007 From: Maciej Swat
This discussion is a perfect example that we need to provide clear explanation what we mean by "standards", "models" etc. In the report when we said "standard" we meant, as Dan pointed out "computer science-based standards for representing models". I know that JSim pushes the meaning of word "standard" to yet higher level, which is a great idea, because, as Jim pointed out we may have standards for exchanging information but the information itself might be not standardized. This is precisely where the need of precise definition what we mean by "standards" comes handy.
During the Model Sharing Mini-Workshop we have spent some time discussing what a "model" means. The word "model" might have different meaning for different scientists.
After that , there came a time when somebody asked question "What is a cell?". And this was a hard question. James Glazier proposed organizing a workshop to answer this question, because, depending on what type of modeling you do your effective definition of a cell will look differently, because your modeling focus will look differently depending on the problem. Since during the workshop we tried to come up with standards for model sharing at the cell level we had to ask questions how we would represent cell in our standards. It is not very trivial and requires talking to different research groups to find consensus definition.
Anyway, this is really great discussion, and I am sure that it will result in more precise set of definitions and this will form grounds for further work on standards themselves.
Since both CellML and SBML can be used to code e.g. models which have imbalanced units, maybe it would make sense to include options in those two ML for users to specify units (perhaps this feature already exists) and then models with correct unit balance would be curated. In any case it looks like we will have to answer another question: "What is a standard?" .
I am curious how this will be resolved. In any case, let's keep this discussion going .....Maciek >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>