Re: [SED-ML-discuss] Stochastic runs of CellML models?

[Herbert S. <hs...@gm...>]

I was thinking of  the situation where the differential equations have
nothing to do with reactions and species, for example someone might encode
the lorenz attractor as three rate rules. It doesn't make much sense to
decompose these into a set of reactions and species even though one could -
I have just to see what they looked like . I've not kept up with SBML but
can one have rate rules that are associated with general state variables
and NOT species, or must all variables be species? This was something I
wanted to be able to distinguish when SBML 1 was developed but it never
became part of the standard. Is it possible now to distinguish species
variables from generic state variables?

I think in CellML there is no way to determine what kind of ODE it is
without some kind of annotation.

I had a look at that paper a while ago but wasn't able to spend long enough
to decipher the theory. Doesn't any one have the paper explained in more
simpler terms? I think I might have understood it has something to do with
determining the Jacobian but not sure?

Herbert

On Sun, Dec 20, 2015 at 6:03 PM, Chris J. Myers <my...@ec...> wrote:

> I don’t think so.  The BioCham method I believe assumes that values that
> are targets of ODEs are species and other values are constants.  You really
> should take a look:
>
> http://www.sciencedirect.com/science/article/pii/S0304397514006197
>
> Chris
>
> On Dec 20, 2015, at 3:53 PM, Herbert Sauro <hs...@gm...> wrote:
>
> Wouldn't you need CellML annotation to identify the differential equations
> that correspond to rate of change of species, assuming they in turn have
> been constructed such that they can be separated?
>
> Herbert
>
> On Sun, Dec 20, 2015 at 2:50 PM, Chris J. Myers <my...@ec...>
> wrote:
>
>> One could use the BioCham approach to extract the reactions from the math
>> like Mike does with the matlab to sbml converter.
>>
>> Chris
>>
>> Sent from my iPhone
>>
>> On Dec 20, 2015, at 2:29 PM, Lucian Smith <luc...@gm...>
>> wrote:
>>
>> That was actually going to be the next part of my question:  stochastic
>> Reactions are a 'standard thing' in SBML, but I believe nobody ever tries
>> to interpret the 'rate rules' in a stochastic manner.  If they did, that
>> could presumably be applied to the CellML side of things.  Conversely, if
>> CellML *did* have some sort of stochastic interpretation, the reverse could
>> be true.
>>
>> For now, it sounds like I should just not try to generate stochastic
>> tests for CellML at all, and stick to reactions for SBML models.  Although
>> if we want to explicitly claim that the SBML RateRule construct is 'immune'
>> to stochastic simulations, I should include that, too.
>>
>> -Lucian
>>
>> On Sun, Dec 20, 2015 at 12:09 PM, David Nickerson <
>> dav...@gm...> wrote:
>>
>>> Hi Lucian,
>>>
>>> Short answer - I'm not aware of any simulator that would know what to
>>> do with a CellML model and a Monte Carlo simulation.
>>>
>>> Long answer - something intelligent sounding about the math being
>>> directly encoded in CellML vs the process description of SBML being
>>> used to generate the math depending on your simulation algorithm (I
>>> don't know if you can do a stochastic simulation of an SBML model that
>>> only consists of math, no reactions?). Andrew Miller did have a
>>> proposal for something along these lines with SED-ML, but requires
>>> extra information provided for the CellML (or SBML) model
>>> (distributions, ranges, etc.).
>>>
>>>
>>> Cheers,
>>> Andre.
>>>
>>>
>>> On Sat, Dec 19, 2015 at 8:04 AM, Lucian Smith <luc...@gm...>
>>> wrote:
>>> > So, SBML has a tradition of allowing the user to encode a single
>>> model, and
>>> > allows you to simulate it deterministically or stochastically.  Does
>>> CellML
>>> > have the same capability?  I'm obviously asking from the perspective of
>>> > SED-ML--if the KiSAO terms for Monte Carlo simulation would mean
>>> anything to
>>> > a CellML simulator.
>>> >
>>> > -Lucian
>>> >
>>> >
>>> ------------------------------------------------------------------------------
>>> >
>>> > _______________________________________________
>>> > SED-ML-discuss mailing list
>>> > SED...@li...
>>> > https://lists.sourceforge.net/lists/listinfo/sed-ml-discuss
>>> >
>>>
>>>
>>>
>>> --
>>>
>>>
>>> David Nickerson
>>> about.me/david.nickerson
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> _______________________________________________
>>> SED-ML-discuss mailing list
>>> SED...@li...
>>> https://lists.sourceforge.net/lists/listinfo/sed-ml-discuss
>>>
>>
>>
>> ------------------------------------------------------------------------------
>>
>> _______________________________________________
>> SED-ML-discuss mailing list
>> SED...@li...
>> https://lists.sourceforge.net/lists/listinfo/sed-ml-discuss
>>
>>
>>
>> ------------------------------------------------------------------------------
>>
>> _______________________________________________
>> SED-ML-discuss mailing list
>> SED...@li...
>> https://lists.sourceforge.net/lists/listinfo/sed-ml-discuss
>>
>>
>
> ------------------------------------------------------------------------------
> _______________________________________________
> SED-ML-discuss mailing list
> SED...@li...
> https://lists.sourceforge.net/lists/listinfo/sed-ml-discuss
>
>
>
>
> ------------------------------------------------------------------------------
>
> _______________________________________________
> SED-ML-discuss mailing list
> SED...@li...
> https://lists.sourceforge.net/lists/listinfo/sed-ml-discuss
>
>