Team:Imperial College London/Lab Diaries/Modelling team


Lab Diaries Overview | Surface Protein Team | XylE Team | Vectors Team | Modelling Team
Here are the technical diaries for our project. We've split them up into three lab teams and the modelling team. We think it's really important that absolutely anyone can find out what we've been doing. For a really detailed look at what we did, and when, you've come to the right place!
Week 4
Day Monday Tuesday Wednesday Thursday Friday
  • Piotr, Anita and Kyasha start work on modelling.
  • Clearly identify all steps of the output pathway and draw out the diagram. Start with the simplest case and build up from there.
  • It was recognised that modelling of output amplification model would be beneficial as model could give answer to how many amplification steps would be most effective in given context.
  • It was decided that other model to be done afterwards would involve modelling the activation of our ComD receptor and the disaply of proteins.
  • It was concluded that the amplification model except for transcription and translation was involving just enzymatic reactions.
  • Tried approaching the output amplification model using Michaelis-Menten kinetics as Kyasha had already some experience in doing that.
  • Decided to have a go on implementing the simple production equations in MatLab.
Week 5

Day Monday Tuesday Wednesday Thursday Friday
  • Attempt implementing 1 step amplifcation into Matlab.
  • Continue developing models/equations/reactions of 2 and 3 step amplification.
  • Try using TinkerCell to implement the 2 and 3 step amplification systems and then export them to SimBiology MatLab interface for verfication.
  • Prepare short slide presentation for advisers for 12.00.
  • Meeting with students from the Royal College of Arts later on during the day.
  • Keep implementing the models in MatLab.
  • Start looking for constants needed.
  • Continue looking for the relevant constants.
  • Meeting with Dr Matthieu Bultelle was arranged.
  • First steps in implementing even simple systems of equations into MatLab are being taken slow.
  • Implementing in TinkerCell was successful. However, its results need to be verified by coding in MatLab later on. We hoped to be able to generate set of equations for MatLab after creating the visual network in TinkerCell, but this functionality was not working in the program.
  • Continue developing models/equations/reactions of 2 and 3 step amplification.
  • Presentation was given
  • We didn't manage to find many information on relveant constants. We decided to arrange meeting with advisers for next day to help us with the issue.
  • We need to look for estimates of the necessary values. It is very unlikely that we will be able to find the exact constants without performing experiments.
  • During the meeting, it was discovered that there is a major error in our model. Our system does not obey Michaelis-Menten kinetics. We have to remodel using the law of Mass Action.

Week 6
Day Monday Tuesday Wednesday Thursday Friday
  • Find the partial differential equations that describe our system according to Law of Mass action
  • Start implementing it into MatLab
  • Since we ended up having 14 differential equations to model 2-step output amplification, attempt implementing the model in TinkerCell as its environment seems to allow easier implementation of that situation.
  • Finish implementing the model into MatLab and TinkerCell
  • Kyasha left to the wet lab team. Anita and Piotr remain in dry lab.
  • Find constants.
  • Find protein production constants and TEV reaction rate constants.
  • We got loads of equations for 2-step amplification model so we did not manage to implement all models
  • Implementation in TinkerCell proved to be quite intuitive and was done.
  • Implementation of codes was completed today.
  • TinkerCell and MatLab generate similar results.
  • We didn't manage to complete the task.
  • The orders of magnitude established - ready to run simulations.
Week 7
Day Monday Tuesday Wednesday Thursday Friday
  • Implementing the constant ranges in the output model.
  • Comparing the results between the models.
  • Start modelling the protein display signalling to find the concentrations.
  • Explain the oscillations that are occuring in the output amplification model.
  • Research stiff differential equations.
  • Research on receptor (especially MAP kinase).
  • Finalise results for amplification model.
  • Prepare presentation.
  • During the meeting it was decided that we should test amplificiation model for its sensitivity to change of parameter values. This is to determine the key parameters to be measured in the lab.
  • Discuss the experiments for amplification module further.
  • Start on modelling the protein display on cell surface.
  • The task is accomplished. However, unexplained oscillations are observed for some specific values.
  • We got rid of the oscillations in our model (by using ode15s instead of ode45).
  • Presentation ready.
  • Reading on stiff equations done.
  • Key sensitive parameters determined.
  • Experiments initially discussed with Chris.
  • Further consult on experiments for determining parameters needed for modelling with James and Wolf.
  • Initial thought for a protein display model implemented in MatLab.
Week 8
Day Monday Tuesday Wednesday Thursday Friday
  • Test the protein display model and find to what kind of values it's sensitive to.
  • Define the control volume around the bacterial cell.
  • Start on testing the model.
  • We realised that it would be worth adding the last step (colour production by dioxygenase) into our amplfication models.
  • We need to finalise testing and formulation of certain assumptions regarding the amplfication (cell death in particular).
  • Consider helping RMIT- Australia in modelling.
  • Keep exploring the prospective solutions to our system going unstable.
  • Try to implement the last bit of amplification in TinkerCell. Maybe it will deal with it well.
  • We couldn't reach the testing as the first model "Display 1" was simulating reactions as if the were taking place inside the bacterial cell while in reality the take place outside. Then we hit an issue of defining the size of volume around the cell. We didn't manage to resolve that problem today.
  • Control volume defined.
  • Testing of model done.
  • Protein model pretty much finalised.
  • The colour production almost complete, however complications were encountered regarding the cell death (colour compound slowly kills cells).
  • Basically we're stuck: scaling didn't help neither did trying idfferent solvers.
  • Meeting with Matthieu gave us some prospective solutions that we can explore.
  • Alternatively we can have a look at SimBiology or MatCont.
  • Helping RMIT: their request is for protein engineering simulation - we cannot do that
  • We know how to use the ''splinetool'' in Matlab.
  • Specifying the time span helps to reduce the simulation time.
  • Model was implemented in TinkerCell.
  • Increasing relative or absolute tolerance in Matlab prevents some solutions from going negative. The key point is that we are only interested in the first few minutes of the experiment as the reaction to catechol is really quick.
Week 9
Day Monday Tuesday Wednesday Thursday Friday
  • Bank Holiday
  • Compare Tinker Cell results against MatLab simulations (Amplficiation model).
  • Compare SimBiology package (in Matlab) results against equations we have derived. (amplification model).
  • See how we could explore the problem of false negatives in Protein Display model.
  • Discuss the protein display false positive.
  • Prepare presentation for the meeting.
  • Compare the catechol colour response (cells alive vs. immediate cell death).
  • Discuss with Matthieu on the issue of negative concentrations, volume approximations made in protein display model and the idea of modelling the false positive case.
  • Update Modelling section of the Wiki.
  • Update list of experiments (especially for Protein Display Model) and discuss with Chris and Wolf.
  • Keep modifying wiki.
  • Keep writing little bits that were left out previously.
  • Tinkercell is not very good at dealing with very high or low values.
  • SimBiology package generates same graphs as our manual simulations. The interface is quite easy to use.
  • Decreasing absolute and relative tolerances is becoming accpeted as a solution to a negative concentrration problem.
  • Modelling of false positive in display protein is postponed as more research need to be done on the affinities etc.
  • Presentation was given - feedback is to be implemented.
  • It was deduced that we should model the cells as alife, because of the slow cell death induced by addition of catechol.
  • Matthieu was consulted and the feedback is posted here.
  • Restructuring modelling wiki is now on!
  • Experiments discussed with Wolf and all should be doable.
  • Wiki updating will continue next week.
Week 10
Day Monday Tuesday Wednesday Thursday Friday
  • Do the referencing of all Modelling Wiki pages.
  • Look into using HTML and Javascript.
  • Make new tables for the Objectives section.
  • Make a jquery accordion list for the first Modelling page.
  • Rewrite pages into html.
  • Update tables (Variables and Constants).
  • Last day!!!
  • Finish rewriting pages into html.
  • Meeting with the advisors at 2pm.
  • Referencing finished.
  • Keep on looking into HTML.
  • New tables completed.
  • List completed.
  • Rewriting into html not yet completed.
  • Tables completed.