Team:Newcastle University/Conclusions

Conclusions
The major outcome of the project is our demonstration that the basic biological function on which the whole project depends - moving two-component quorum-sensing systems from one strain of Bacillus to another strain can be achieved, and the two-component system can continue to function as it does in the original strain. We have decoupled the production and detection of quorum sensing peptides. Although we didn't achieve our overall goal of evolving an ANN and implementing it in B. subtilis, this a major step towards the achievement of out original goals.

The preliminary work involved identifying suitable target pathogens, the quorum communication peptides produced by them, and the amount of overlap between peptides produced by different species. This took much longer than anticipated, and involved a lot of searching the literature and online databases, but resulted in a database of species and peptides which is a good training set for the system.

Software development also took up a lot of our time, since all of the code had to be developed from scratch, and this could not be done without learning a lot of skills, including the principles of neural networks and evolutionary algorithms, advanced Java, CellML and JSim, databases and JDBC and Web services. The fact that the software works and the different parts communicate with each other is very rewarding, and the software may be a valuable resource for future iGEM teams, if they choose to work in this area.

The lab work was also a positive experience, as we managed to clone the proof of concept BioBrick into an integration vector, and get it to integrate into the chromosome of the Bacillus. Once again, this required the acquisition of a host of new skills, both at the bench and microscopy, flow cytometry and, for some of us, computing and the use of Wikis.

Overall, the 2008 Newcastle iGEM team had a lot of fun, learned a lot of skills, and made (hopefully useful) contributions to the BioBricks Repository, as well as local databases and software tools. Jamboree, here we come!

Future Work

 * The Bacillus chromosomal integration vector we used is not BioBrick compatible. It may be possible to modify it so that it is, or provide simple and reliable protocols to transfer a BioBrick part from compatible vectors into this one.
 * Need to further characterize the subtilin Brick part. We started to do this with the flow cytometry and microscopy, but more data-points would make simulations more robust.
 * Extend to the full original plan. We have demonstrated that extra-cellular quorum peptides can be sensed and that they can control the expression of reporter genes. The next step would be to implement a stripped-down bacterial neural network, and finally to try out one of the full solutions produced by the evolutionary computation.