Modelling Notebook

Week 1 08/06/09 - 12/06/09

We split into pairs and began writing a system of ODEs for each of our three possible wiring diagrams. The aim of this was to decide which genetic circuitry would work best, depending on how the models performed. In order to get any data out the models however we had to ask the biologists on the team to help us find parameters, although this proved to be challenging.
Started to program our ODE's into our RK4 program (written in C)

It became apparent, from the modeling, that the third circuit design was superiour to the other two for several reasons. As such the biologists could start thinking about starting their lab work.
Monte Carlo simulations were written in C to test the viability and robustness of our model

First proper chemotaxis model’s created without animation, and could only handle a few bacteria at a time. Consideration went into more practical sides of the chemotaxis, such as average time to travel a distance. Bacteria respond to a concentration, as opposed to distance. This was done in Matlab.
ODEs were refined and internal dynamics of the cell were studied further.
PDEs were investigated as a possible way to integrate several components of models

Program written to estimate the point at which quorum sensing would turn on by simplifying system to steady state bacteria each producing HSL in a control volume of water.
Equations were added to the ODEs to describe the transcription of mRNA molecules.
PDEs turned out be too difficult to model, as a result the chemotaxis model was converted into C so that it would be possible to merge both the chemotaxis and internal dynamics into a single working model.

The program simulating quorum sensing level is incorporated into a population model so that population grows, and with this HSL, until lysis occurs and population decreases. Time for lysis must be less than the replicating time or else the population will not die. (although growth may stop when quorum sensing turns on, as protein production will inhibit growth).
ODEs were revised after input from the biologists

SimBiology is first played with, but was difficult to customise to our needs and so work was continued with the chemotaxis model in Matlab. The model is now for bacteria contained in a pipe and moving in a changing chemical gradient towards a source in the form of a crack.
ODEs were revised again after consultation with the biologists.

A stochastic simulation was done of the deterministic model, using the tau-leap method. This was programed in C.
Further work was done on combining chemotaxis and internal dynamics in C.
The first working SimBiology model is created and is exactly parallel to the C differential equation model apart from being modelled by mass action equations, several of which are governed by hill functions.
Work went into changing the chemotaxis model so that it can output video and run for thousands of bacteria.

Critical radius and critical density for quorum sensing was invesitgated.
Quorum sensin system was analysed.
Writing of the wiki was started
SimBiology model was made more complex due to more accurate and in depth knowledge about the lac operon. The new complex model also works well.

Lots of work done on the wiki!
The chemotaxis model video outputs are now completed and are in full colour. Work had to be done to convert GIF format into a useable one for the wiki. We also had fun making things like smiley faces that turn into an iGEM logo by playing with colonies of bacteria attracted each to different point sources.