Team:Evry/ODE model
From 2012.igem.org
Model using Ordinary Differential Equations (ODE)
Overview

This global model represents the tadpole as a three compartment system:
 The skin that produces (or receives) auxins, denoted S in the equations;
 The blood that transport auxins to the organs, denoted B;
 The organs (called receptors) that interacts with auxin molecules, denoted R.
 The auxin concentration inside a compartment is homogeneous This condition is inherent to this kind of modeling.
 All the auxins that enter the receptor come from blood. Considering blood speed and the big density of capillaries, we assumed that the auxin coming from other tissues can be neglected. This assumption will remain while the receptor isn't close to skin.
 The auxin flow follows the concentration gradient between compartments. This assumption is based on the fact that a small capillary is nothing more than a hole in a cell layer. So the exchanges between the surrounding cells and the capillary are mutual.
 Permeabilities are symetric. We have considered that for 2 issues X and Y the permeability from X to Y is equal to the one from Y to X.
 S in m^{2}, represents the area of the exchange surface between the two compartments.
 P in m^{2}, represents the permeability of the membrane between the specified compartments.
 C in [quantity] / m^{3}, represents the concentration of auxin in the specified compartment
 For permeabilities:
Playing with skin permeability influences highly the auxin quantity in each compartment. This is expected since the skin is the emitter in our system. As for receptor permeability, the higher it is the less we get auxin in skin, leeding us to a conservative behaviour in auxin quantity between skin and receptor compartments. Tuning the blood permeability seems to be less significative for auxin quantity in the compartments giving that the blood plays a transport role.
 For die rate:
Tuning the degradation rate has leed us to a signficant value 0.00001 mol after which, we get a different behaviour of the auxin quantity evolution in receptor. As it represents the destination compartment of auxin emitted, it may explain why it was the most affected.
 For Surfaces:
Simulations with different surfaces values do not really show big differences in the main landscape which keep the emitter with the higher quantity of auxin. Tuning the skin surface gives a proportional evolution of auxin quantity in receptor and in emitter, in the other hand tuning the receptor surface produces an antiproportional evolution of auxin quantity in the two compartments.
 For Volumes:
The auxin quantity in each compartment seems to be sensitive to the volume of the latter, except for the receptor. The system is not sensitive to small variations of the receptor volume.

This first graph is obtained considering no auxin in the tadpole and c_{w}=5.10^{3}mol.L^{1} which is the concentration we wanted to use for experiences. The result shows that the quantity of auxin in skin and in the receptor grow until a limit value.
If there has been injected auxin in the receptor and in skin: n_{s}(t=0)=10 mmol n_{r}(t=0)=10 mmol, we can observe a decrease of auxin in receptor before an establishment of the equilibrium.
Here , the concentration in water has been devided by ten c_{w}=0.5.10^{3}mol.L^{1}, while the initial value of skin is still n_{s}(t=0)=10 mmol. We observe first a big increase in the receptor due to the diffusion of the auxin intially injected in skin then a decrease because of the small water concentration in auxin.
Using this very idealized view of the tadpoles makes it possible to model it simply : The processes happening in the system can be approximated using Ordinary Differential Equations (ODE), one of the simplest form of differential equations; Plus, the organs repartition and shape are not taken into account. As showed in this page, these equations can be seen as approximations of more realistic PDEs where space considerations have been removed.
Although our model is very general and simple, it allows to have a quantitative view. As many parameters would require much experiment to be estimated correctly. At this stage this quantitative model only allows us to make some prediction about the global behaviour of the system and gives us hints about the reasons for the success or failure of experiments.
Assumptions
There are the different assumptions we were constrained to make in order to model the system:Model description
Equations
Each compartment is modeled by a differential equation representing the evolution of the auxin quantity as a function of the time.
Each equation is composed of two kinds of terms: synthesis or arrival (called born) and degradation (called die). The creation term can represent either a creation of auxin in the compartment or an arriving quantity of auxin from another one. The following equations represent the case where auxin flows in from water. In the case where we consider auxin being synthesized directly in skin, this term would be the mean rate of synthesis. This rate being computed integrating data from the plasmid repartition model and the auxin production model, as described in our derivation from PDEs. In the same way, the degradation term can either represent a natural degradation of molecules or a quantity leaving the compartment.
In this system,D_{die} is a degradation rate, the other D terms represent the flows between the different compartments. We made them depend on the concentrations of both the in and out compartments as explained in assumption 2. Their mathematical formulation is the following:
Where:
These flow equations are based on Newton's law of cooling where the difference between the concentrations of the two compartments gives the direction and magnitude of the flow. This allows us to model in a single equations the two opposite flows between the compartments.
Parameters
The parameters of the model are hard to estimate because direct experiences concerning auxin's behaviour in the xenopus where never made before, as it is a plant hormone.Moreover,the parameters are an aggregation of different physical values. This is due to the high level view provided by the model. The different parameters are:
Description  Symbol  Type  Values  Experiment Proposal 

Permeabilities  P  similar molecule in litterature  here  link 
Contact surfaces  S  calculated  here   
Volumes  V  calculated  here   
Degradation rate  D_{die}  unknown  unknown  link 
Creation rate  D_{born}  computed  our derivation from PDEs   
Sensitivity Analysis
There are 2 kinds of parameters in this model; those concerning the geometry of the tadpole and those concerning the behaviour of auxin in the tadpole's tissues. As auxin is a plant hormone, we had to evaluate these parameters by taking those of similar molecules. That's why the uncertainties on these measures are very big. For this category, we observed variations from reference value (ref) from ref*10^{3} to ref*10^{3}. You can see the effect of these variations:
Results
By modifying the initial conditions and c_{w} we have observed globaly 3 different behaviours of the system:Conclusion
This model being simplified to the maximum, has let us, thanks to an estimation of very few parameters, to see the global behaviour of the system. It helps us to determine expected behaviours and helps us to establish possible reasons of dysfunctions. Nevertheless, quantitatively, this model is too imprecise because it doesn't take tadpole's geomoetry into account. This is the aim of our most elaborated model which will use partial equations.
You can find the model here.References
Other possible topologies
With auxin in the external medium:With a specific receptor organ: