<< click on the image to enlarge!
Example of region concept using an unstructured grid. FEīs are tagged with region IDīs which can be used subsequently to define regional properties of a model.
:: Flexibility: Mixing Ionic Models in a Domain
LIMPET is designed to set active tissue properties in a domain on a per region base. In a finite element mesh, each element may be tagged. Such tags or arbitrary sets of tags can be used to define a region. To every region an ionic models can be assigned, each ionic model can be augmented with any of the available plugins, and almost any default parameter in any ionic model or plugin can be overruled. A negotiation layer between ionic model an plugin is required to determine whether an ionic model provides proper support for a plugin. For instance, the plugin IA, a hypothetical potassium current that activates at large positive polarizations, influences the potassium concentration of the ionic model. For this sake, the plugin has to negotiate first whether the ionic model keeps track of potassium or not. If so, the plugin needs to know how to update the potassium concentration of the ionic model. A complex example showing most of the LIMPET features is shown here:

num_regions = 3
region[0].description = "ENDO"
region[0].im = "UCLA_RAB"
region[0].im_param = "gNa+10%:gKs*0.95"
region[0].im_sv_init = "sCpacing_bcl_100x300ms"
region[0].im_sv_dumps = "cp:cs:ca_i"
region[0].plugins = "IA:EP_RS:Muscon:SAC_KS"
region[0].plug_param = ":q=2.71,N0=7.5e4:::"
region[0].plug_sv_dumps = ":n:olambda,z,Q1,Q2:"
region[0].num_IDs = 4
region[0].ID[0] = 1
region[0].ID[1] = 2
region[0].ID[2] = 3
region[0].ID[3] = 4

region[1].description = "EPI"
region[1].im = "PUG"

This shows a practical example that demonstrates the flexibility of LIMPET. The intracellular grid is subdivided into 3 regions (specified by num_regions). Shown here is the con- figuration of the first region only. We name the region ENDO and assign the UCLA_RAB model. We also would like to overrule some of the default settings of the model (see reference xx). Here we increase the sodium conductivity by 10 percent and we reduce gKs by 5 percent. Instead of using the default initialization of state variables we use a different initial state vector which we have generate with the single cell code bench where we paced a single cell 100 times at a basic cycle length of 300 ms. Furthermore, besides Vm and φe which are output by default we also output dyadic space, (cp) the subsacrolemnal space and (Ca_i) the cytosol. Several plugins are switched on, EP_RS (electroporation with resealing), IA (hypothetical potassium current that activates at large depolarizations), MusCon (the Hunter - McCulloch - terKeurs model to compute active mechanical parameters) and SAC_KS (model of stretch activated channels). Several default parameters of the electroporation models are overruled, whereas all other plugins use the default values. We also output several state variables in the plugins including the number of pores n formed due to electroporation and several active mechanical parameters. The region to which we apply all these settings is composed of 4 subregions where the subregions are formed by finite elements that are either tagged with the IDs 1, 2, 3 or 4.
Calcium concentration in different cell components such as the dyadic.
[HETEROGENEITIES] . previous page . next page .