System Simulation

Introduction

In order to verify the availability and probable effect of our design scheme, simulation mathematical models based on Cellular Automata technique have been conducted.

In catering to our past experiment scheme, we have constructed two models as follows:

• Light Sensing Model: use light to trigger the asymmetry process of the colony;
• Movement Model: use more complicated pathways to induce the break of symmetry.

A cellular automaton is in nature a finite-state machine in discrete time as well as space studied in computability theory,mathematics, physics, complexity science, theoretical biology and microstructure modeling.

This model provides a significant reference to the appraise of realistic experiments by simulating the whole pattern changing process, which consists of division, movement, death and some relevant ones.

Light Induced Model

Micro Model

To display the simulation results directly, we have constructed a micro model based on cellular automata. Each minimum spatial unit is able to hold up at most one cell, which can be characterized by a cellular automaton; Meanwhile, another property that wonders in every minimum spatial unit is the density of AHL, which relies on the density diffusion equation. The detailed cellular automata rules would be as follows.

a) Cell types

Each minimum spatial unit in this cellular automata model is able to hold up at most one cell. Cells can be divided into two types, normal cells(marked by green) and special cells(marked by red). Red cells are transferred from green cells by the trigger of light. The type of cells is denoted by ‘C’ as follows:

‘n’ represents the n th period, ‘(i, j)’ represents the location of the cell. Same below Red cells emit AHL at the rate of ‘q’ in every period.

b) Cell Division and Cellular States

Both kinds of cells are likely to divide. A mature cell possesses a division probability of Pdiv every time unit and gets two green cells after division. The newly generated cell would grow closely to the original cell and would emerge at the eight neighboring positions with equal probability. If the eight neighboring positions had been occupied, then the original cell would not be able to divide.

Consider the realistic condition where continuous cell division is impossible, so we use Φ to denote cellular states. At the very beginning Φ=1; Then with the cells growing, Φ=Φ+1 with every period; When Φ=τ, the cell would be mature enough to divide with a division probability of Pdiv during every period; After every division, the Φ(n;i,j) of the original and new cells would both change into:

However, in our system, as a signal molecule, AHL can trigger the antisense Ftsz part which can be functional as a division repressor. So Pdiv is actually a constant but a parameter proportional to the density of AHL. With several steps of calculations and experiments, the relationship between Pdiv and the density of AHL can be induced as follows:

Among it both ‘a’ and ‘b’ are unknown parameters.

c) Cell Movement

In the light sensing model, due to the big influence of light from the outer environment, cell movement has almost nothing to do with final results both in simulation or realistic experiments. As a result, we simplify the movement of cells. However, the movement of cells would no longer be negligible in the afterwards models. Further details on its rules and results to come afterwards.

d) Cell Death

In all models mentioned in this text, cell death has merely minor influence on our final results so it is neglected in all.

AHL Density Equation

AHL density can be expressed as u(r,t). Take the discrete characteristic of cellular automata into consideration, just denote it as . As a solute, AHL has the phenomenon of diffusion and decomposition. At mentioned above, red cells would spread out AHL. In conclusion, AHL complies to diffusion equation:

• u: AHL density; D: diffusion constant; q: AHL releasing rate of a single triggered cell; γ describes the decomposition process of AHL; m: total number of cells.

Simulation Procedure

At the very beginning, AHL density u(r, 0)=0 on the infinite plane. In order to eliminate the asymmetry in the results that might be caused by the asymmetry of the original state, there is only one green cell at first. Choose several separate cells manually to shed light on. With the sliding improvement of the colony from merely one cell, green cells would eventually reach the light area where they would transform into red cells. Red cells begin to release AHL. When the quantity of red cells accumulates to a certain threshold, high density of AHL in certain areas would repress the division of cells and begin to trigger the differentiation of cell and therefore break the symmetry of the colony.

Results

Macro Model

The micro model is somehow quite easy to understand. We can relatively directly observe and analyze the location and state of each cell, so that this model is suitable for small-scale colony pattern vicissitude analysis. However the application of several random factors like division probability or others eventually complicates the anticipation process. Moreover, variations exist every time of simulation. We have to observe the shape of colony directly(with eyes) on a larger scale, which would obey the statistical rules. For the sake of seeking the most probable situation, we have a macro model to simulate. Macro model differs from the micro model mainly in that the number of cells in a spatial unit no longer limits to one but pertains to another characteristic: cellular density. The macro model would on one hand reduce the quantity of calculation and on the other hand reflects the possible results more realisticly.

Compared to micro model, the AHL density u(r, t) for every spatial unit possesses the same meaning and remain compliant to equation(4). Whereas, the characteristic of cells is no longer weighed by whether existed or not, nor does the type or state of a single cell be considered, but using the density of green normal cells and that of red special cells

The Cellular Density Logistic Model

In the micro model, we can discuss the division process and growing period of single cells, but not in the macro model any more. For the concept of single cells is eliminated and instead we just take into consideration the quantity of cells in certain areas which would obey some kinds of rules. Here, we applied the classic Logistic Model to lineate the quantity visisitude of colony. With the original number of cells ranging from 1 and 200 and a environmental limit of 100, the changing situation of the number of cells is shown as below:

• Logistic model with initial cell number of 1 and 200

The equation is shown as follows:

[[File:Seuofun5.GIF|center]

Among which, k is a proportion constant, Nm denotes the environmental cell volume at the location r and time t and is a function of the AHL density u(r, t); N(r, t) denotes the number of cells at the location r and time t.

Logistic Model is a somehow realistic model which takes not only the growth of number but also the environmental tension after the growing of the number into consideration and therefore makes it more practical than the miro model.

As high AHL density would repress cell division, we can regard the environmental volume as a function of AHL density. AHL density gets higher and environmental volume Nm gets lower. The relationship might be as follows:

[[File:Seuofun6.GIF|center]

Among which both 'd' and 'e' are unknown parameters,and d>0.

After the change of cellular numbers, due to the background that red special cells can only be transformed by light, all newly-built cells in this Logistic Model would be green cells.

b) Cell Movement

As the movement of green cells is random movement, therefore the movement of green normal cells relies on the formula below:

Among which 'V' denotes the diffusion constant or a reflection of cellular rate.

The movement of red cells is the same.

Other parts and simulation principles of macro model are just the same with the micro model.

Results

Division Inhibition