Chronological Thread 

< Chronological >      < Thread >

---

Some comments from Big Brother ;-)

"# 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"

Why is-that? As far as I know, there is nothing in SED-ML (the new name of
the format that try to implement some of MIASE support) that is specific to
SBML. If you disagree, we would be happy to get in touch to solve that.
SED-ML is really meant to be entirely format independent ... as far as the
format is some kind of XML, since the variable are targeted using XPATH
(and even that could be argued, since an XPATH can be transform into any
kind of pointer)

"# 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. "

Probably because we overlooked the comment. It is of course not true. Any
dataGenerator can be used on x, y or z.


"MIASE also lumps graphing and simulation together, which we do not believe
to be appropriate for CellML"

Yes ... and no. SED-ML says what to plot against what. Not how (except for
the linear vs log).

"# 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"

I completely understand that. We are currently aware of five
implementations under way, by COPASI, VCell, BioUML, SBToolBox and SBW.

But I think there is a slight misunderstanding here, that is entirely our
fault. Initially well separated, the MI and the format tended to blur
around the new year. We are doing our best to more cleanly separate both.
SED-ML is akin to SBML and CellML. MIASE is akin to MIAME. CellML does not
have to implement SED-ML. This is the job of the simulation software.
SED-ML will call CellML and SBML models and describe what to do with them.

MIASE is another issue. It is meant to describe the minimal information to
encode in order to reproduce a simulation. CellML graphing and simulation
metadata are/will be MIASE compliant. Otherwise they are not sufficient.

But in fact MIASE is for the moment just an idea. And we will be in touch
with everyone in order to formalise it more, and transform it into a
"MIBBI-grade" checklist (MIBBI is the portal that federates the MIs.
Publication to appear inn Nat Biotech next week)

Cheers,

James Lawson wrote:
> Dear CellML community,
>
> The ABI CellML team meeting minutes for last Wednesday (2008-07-30) are
> now available at:
> http://www.cellml.org/meeting_minutes/abi-meeting-minutes-2008-07-30/
>
> Kind regards,
> James Lawson
>
> _______________________________________________
> cellml-discussion mailing list
> cellml-discussion at cellml.org
> http://www.cellml.org/mailman/listinfo/cellml-discussion

---

• [cellml-discussion] ABI CellML team meeting minutes 2008-07-30, James Lawson, 08/05/2008
  • [cellml-discussion] ABI CellML team meeting minutes 2008-07-30, Nicolas Le novère, 08/06/2008
    • [cellml-discussion] ABI CellML team meeting minutes 2008-07-30, Dagmar Köhn, 08/08/2008

---

Archive powered by MHonArc 2.6.18.