Menu
▾
▴
Re: [sbml-dynamic] SBML grad student
|
From: Chris J. M. <my...@ec...> - 2013-07-12 22:28:29
|
Was there an attachment? If so, I did not get it. What you described though is a bit different than the approach that we had in mind. Basically, the event is local to the submodel. This makes it clear what it means when the event fires a destruction event. Namely, it destroys itself. A creation event duplicates itself. This approach is closer to how a cell works actually, so it makes sense to us at least. Chris On Jul 12, 2013, at 11:41 AM, Lucian Smith <lp...@sp...> wrote: > In the interest of sparking conversation, here's an extended Event UML I > threw together along these lines. The basic idea would be that a > 'Creation' would create a new instance of the referenced submodel, and a > 'Destruction' would remove an instance of the referenced submodel. > *Which* instance would have to be determined somehow--I have a vague > idea that it'd be 'the parent of the particular item in the Trigger that > caused the Event to fire in the first place', but I'm not exactly sure > how that'd work, or if it'd work very well. Where the information is, > is an interesting question. > > -Lucian > > * Lucian Smith <lp...@sp...> [2013-07-12 18:12] writes: >> No, there's no document yet. The general idea behind 'dyn' is up at >> >> http://sbml.org/Community/Wiki/SBML_Level_3_Proposals/Dynamic_Structures >> >> but it is very very vague. >> >> Here is my current understanding of what people are thinking so far: >> >> * Creation and destruction will be tied to the 'Event' SBML core >> construct. >> >> * This event result will be 'rule-based', i.e. will describe what to do >> when the event triggers. >> >> * The simplest way to group SBML structures for creation and destruction >> is probably through the use of Submodels from the Hierarchical >> Composition Package. >> >> Am I right that that's the basics? Does anyone see a problem with that >> approach, or a case that couldn't be covered? >> >> -Lucian >> >> * Antoniotti Marco <ant...@di...> [2013-07-12 17:57] writes: >>> Hi >>> >>> tying to get up to speed and getting more involved? is there any document r.e. SBML dyn? I was not able to get to a doc from the web page. >>> >>> Thanks >>> >>> Marco >>> >>> >>> >>> On Jul 11, 2013, at 18:39 , Harold Gomez <Har...@gm...> wrote: >>> >>>> Hi everyone, >>>> >>>> A few comments: >>>> ? "Dynamic flattening" of running models at given stages of simulation seems quite useful and worth exploring. Our group is definitely interested as it would us to share and save the state of model elements after a given number of iterations. What is your intuition in terms of whether this is something that (1) the community would be interested in (2) is likely to form part of the current development of the grammar for the dyn package. It sounds to me as if this would be useful in the context of the end goal (multicellular modeling), but I not sure whether such feature belongs in our current "development cycle". >>>> ? I agree that a better approach to this problem is focusing on the grammar needed to encode the mechanism for signaling dynamic events (addition/removal of submodels) in SBML, as opposed to trying to describe these dynamic biological events themselves. >>>> ? In terms of figuring out what elements should be encoded for sharing and which should be left out to be handled by the local simulator, I have already started making strides with the guidance of Dr. Myers and Keating. I've looked into how this is done by iBioSim, which is 1 of the 2 implementations needed that will use the constructs defined in the dyn package -the other one being Daphne, our in-house simulation tool. iBioSim uses special annotated events to signal to the parser that a dynamic event is taking place, and I think we could definitely improve on this approach by using the groups package, which I am open to help on. I'd like to understand what happens behind the scenes in iBiosim once such models are read, and I think looking at the source code or having a conversation with Dr. Myers or Leandro, his student, may help here. >>>> ? On a related note, as Daphne currently provides no SBML support, I will start implementing core L3 SBML features with the help of libSBML. >>>> I hope you enjoy both the conference and your vacation professor Myers. Please feel free to suggest a time and date to schedule our skype meeting at your convenience, which I hope other members of this work group will be able to join in. Please feel free to suggest ideas or approaches. >>>> >>>> Cheers, >>>> >>>> On Tue, Jul 9, 2013 at 5:52 PM, Lucian Smith <lp...@sp...> wrote: >>>> * Chris J. Myers <my...@ec...> [2013-07-09 22:34] writes: >>>>> Hi, >>>>> >>>>> Sorry for the delay. I'm in London at SB6.0. Putting my reply on the dynamic list. Not sure how many people are on this list beyond Lucian and myself but let's see. >>>> >>>> Looks like there are 21 people on the list. It's just that the two of >>>> us are by far the most wordy ;-) >>>> >>>>>> Thank you for your response. I am trying to get my head around the core of the problem, where we stand, and what's to be done, so please correct me if I am mistaken. >>>>> >>>>> One caveat to my answers is that our approach is at this point simply our approach and not really a proposal per se. The dyn package has not been active, so we just came up with an approach for our needs. Therefore, I'm very open to suggestions for ways to improve as our approach is very simplistic. >>>>>> In the case of a division event, the duplication of the cell compartment happens during simulation and this "duplicated" state is never encoded in SBML. SBML is only used to encode the initial state of the model while denoting which blocks of code contain elements that are to be treated as dynamic for the appropriate software tool (iBioSim) to parse and process such structures accordingly. >>>>> That is correct. I suppose it may be possible to generate the new SBML after the division event, if one desired, assuming we have clear rules as to what the state of the model becomes after the division. >>>> >>>> That's an interesting idea--a sort of 'dynamic flattening' algorithm >>>> that could export a static SBML model for any given point in the >>>> simulation. Hmmm... >>>> >>>>>> What we are striving for in our dyn extension is a set of data constructs (as defined in an XML schema file) needed to standardize the way that different software implementations encode dynamic biological processes. >>>>> That is correct. I think it may be very difficult to encode in a standard way exactly how the dynamic model is to be simulated but we could encode the mechanism to determine when the dynamic event should happen. Multi-cellular simulators may evolve differently after such an event perhaps. There is a lot of diversity in multi-cellular simulators. >>>> >>>> Figuring out what bits to exchange and what bits to leave up to the >>>> local simulator is going to be the trick here, I think. The question >>>> becomes: at what point does it stop being 'the same' model that you are >>>> exchanging? >>>> >>>>>> In a part of the proposal page for the dyn package is stated that " Dynamic structures are usually associated with compound structures such as lists, sets, and arrays". However, given that there is an array and set package specification currently under development, how would our approach to tackling our problem be different from this? >>>>> That statement is a bit out-of-date. There is now a "static" arrays package but there was a proposal for a dynamic arrays package before. The problem with using dynamic arrays is that when you have a death event, you would have gaps in the arrays which is awkward. It might be possible though to do something with "sets". We have a package called "groups" which I think is nearly a "sets" package though would need to be updated to support set math. >>>>> >>>>> I'm not sure if we want to have a general dynamic package that allows creation or deletion of any object. I think one that creates new (or destroys) instances of submodels may be sufficient. >>>> >>>> Restricting dynamic to work on submodels directly definitely seems like >>>> a good way to encapsulate the complexity to one thing while still being >>>> open to a wide range of options. >>>> >>>> -Lucian >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> See everything from the browser to the database with AppDynamics >>>> Get end-to-end visibility with application monitoring from AppDynamics >>>> Isolate bottlenecks and diagnose root cause in seconds. >>>> Start your free trial of AppDynamics Pro today! >>>> http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk >>>> _______________________________________________ >>>> sbml-dynamic mailing list >>>> sbm...@li... >>>> https://lists.sourceforge.net/lists/listinfo/sbml-dynamic >>>> >>>> >>>> >>>> -- >>>> Harold Gómez >>>> Bioinformatics PhD. Student, Boston University >>>> Graduate School of Arts & Sciences >>>> ------------------------------------------------------------------------------ >>>> See everything from the browser to the database with AppDynamics >>>> Get end-to-end visibility with application monitoring from AppDynamics >>>> Isolate bottlenecks and diagnose root cause in seconds. >>>> Start your free trial of AppDynamics Pro today! >>>> http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk_______________________________________________ >>>> sbml-dynamic mailing list >>>> sbm...@li... >>>> https://lists.sourceforge.net/lists/listinfo/sbml-dynamic >>> >>> -- >>> Marco Antoniotti, Associate Professor tel. +39 - 02 64 48 79 01 >>> DISCo, Università Milano Bicocca U14 2043 http://bimib.disco.unimib.it >>> Viale Sarca 336 >>> I-20126 Milan (MI) ITALY >>> >>> Please note that I am not checking my Spam-box anymore. >>> Please do not forward this email without asking me first. >>> >>> >>> >>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> See everything from the browser to the database with AppDynamics >>> Get end-to-end visibility with application monitoring from AppDynamics >>> Isolate bottlenecks and diagnose root cause in seconds. >>> Start your free trial of AppDynamics Pro today! >>> http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk >>> _______________________________________________ >>> sbml-dynamic mailing list >>> sbm...@li... >>> https://lists.sourceforge.net/lists/listinfo/sbml-dynamic >> >> ------------------------------------------------------------------------------ >> See everything from the browser to the database with AppDynamics >> Get end-to-end visibility with application monitoring from AppDynamics >> Isolate bottlenecks and diagnose root cause in seconds. >> Start your free trial of AppDynamics Pro today! >> http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk >> _______________________________________________ >> sbml-dynamic mailing list >> sbm...@li... >> https://lists.sourceforge.net/lists/listinfo/sbml-dynamic > > ------------------------------------------------------------------------------ > See everything from the browser to the database with AppDynamics > Get end-to-end visibility with application monitoring from AppDynamics > Isolate bottlenecks and diagnose root cause in seconds. > Start your free trial of AppDynamics Pro today! > http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk > _______________________________________________ > sbml-dynamic mailing list > sbm...@li... > https://lists.sourceforge.net/lists/listinfo/sbml-dynamic |
