From: Ally H. <a....@ed...> - 2015-08-28 14:02:52
|
Hi, I am currently exploring how a digital Arabidopsis plant model [1] can be rendered in SBML. Looking at the various packages, the Hierarchical Model Composition and Dynamic Processes packages look very useful for capturing the functionality of the model. Having read these specification/drafts I have a few questions that hopefully somebody on this help me with or at least point me in the correct direction. The plant model has dynamic sub-mobels for leaves. When it is time to create a new leaf then a new sub-model is create for this leaf. It would be great to use the Hierarchical Model Composition for the leaf sub-model and use Dynamic Processes for the creation of these sub-models when required. My first question is whether you see this type of thing as something the Dynamic Processes package is attempting to support? Having read the draft specification the type of dynamic functionality supported is tied in very closely to the Cell Behavior Ontology (CBO). While I can certainly see the need for dynamic functionality such as that proposed in the draft I was expecting to see also more generic dynamic functionality such as to simply create a new sub-model instance. Given this tight coupling to the CBO am wondering if my use case if something you feel the package is trying to support? Assuming for the moment that the package should cover the new-leaf dynamics then my next question is regarding how Dynamic Processes would work with Hierarchical Model Composition. In particular, I would like to be able to define variables that are aggregations over the dynamically created sub-models. For example, it would be good to have a variable that is the sum of the masses of each individual leaf. Hierarchical Model Composition does not define such aggregate functions. I assume the reason for this is because there is not great need for them in a static model as one can simply implement the aggregation by explicitly summing up all the static sub-model variables. However, when sub-models are created dynamically one starts to need such functionality. Therefore, while the such details as probably not part of the Dynamic Processes role there is a strong need for other packages to understand the potential impact on dynamic functionality on them. How would one go about supporting aggregate functions of dynamic sub-models? That is about as far as I've got in my current thinking. I have no doubt that those in the community have probably discussed such issues for many years now and I'm sure I'm missed many details that have shaped the thinking. Can somebody possibly advise as to how they see SBML being extended to support my use case or whether such functionality is simply not on the radar? I am planning to attend the Combine meeting in October and have some time to work on these issues and help where possible so I'm keen to know if anybody is working along these lines so I could maybe collaborate with them. Thank you for your time. Ally [1] http://www.pnas.org/content/111/39/E4127 Ally Hume Software Architect, EPCC and Systems Biology modeller at SynthSys The University of Edinburgh Skype: ally.hume -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. |