A large amount of progress has been obtained in the research line of arsenic accumulation, however several other interesting metals (e.g. copper, zinc) had to be abandoned which might provide more interesting applications. For this to occur, the constructs that were abandoned along the way (e.g. SmtA, HmtA) should be finished and characterized. But also for the existing arsenic devices more characterizing can be performed, e.g. more detailed characterization of the uptake capacity of the transporter (using higher metal concentrations in death assays). In this project fermentation was used to supply the culture with medium saturated with air in higher amount than can be attained in shake cultures. This was to ensure a higher amount of gas to be available to enter the gas vesicles, for a future project it might be of interest to sparge nitrogen into the broth to obtain anaerobic conditions. Anaerobic condition will influence the oxidation state of arsenic, probably having an effect on the uptake efficiency. A nice addition to the gas vesicle cluster would be the removal of the 10 time repeat, which was attempted in our project without success.



To make it possible to have interactive graphs and calculators on our Wiki, based on our model, we have implemented our entire model in JavaScript. This included developing our own ODE solver (among other things). We feel that having our model accessible on the Wiki has many advantages and it would be very useful to develop a library to handle ODE (and perhaps stochastic) models using JavaScript. Such a library could support:

  • Graphs that work well in pretty much any browser, including a legend (and possibly 3D functionality). Linked graphs would be a plus.
  • Reading from various sources, like an HTML table, Google Docs, csv files, etc.
  • Integration of ODE models defined by a function that returns the gradient based on the current values.
  • Equilibrium computations. Preferably based on the same function used for integrating ODE models.
  • Fitting a model to experimental data.

"Dumbed-down" user interface

Normally a model is used only "internally" to, for example, predict results and/or analyse them. We have made our model a more integral part of our project by making (parts of) it available right next to our theoretical texts and results, but this could be taken much further. It would be interesting to make an interface to our model that is suited for someone without an extensive background in molecular biology and/or modelling, with easy-to-use sliders and clear visual representations.

A "dumbed-down" user interface would likely not be aimed at doing any of the normal modelling tasks, but would be more of an educational tool, allowing an interactive exploration of a project. This could be immensely useful in giving people insight into an iGEM project like our own.


Our determination of the import rate of arsenic is reasonably accurate, but many more parameters of which we are much less certain remain. On our characterization page several different ways to find these parameters are discussed.