TR: How is your approach different from traditional molecular biology techniques?

TG: People have been modifying genetic systems for years. But, for the most part, it's a trial-and-error approach. They tweak something and see what happens. We wanted to bring a systems level perspective, so we could approach the problem like an engineer. In order to do that, we had to know more about the existing circuitry, so we began to do genetic mapping.

We've been focusing on mapping regulatory circuits [a network of genes that control the chemical reactions taking place in the cell]. If you're trying to figure out the circuitry of a house, you go to the circuit breaker and flip circuits on and off, looking for the circuit that controls the bathroom or the kitchen. We do a similar thing in bacteria, but it's a bit messier. We stress the bacteria in different ways, with different chemicals or extreme temperatures, and then see how each gene responds. If you do this hundreds of times, you can look for genes that change together. For example, if you see different genes whose expression changes the same way under different conditions, we can infer those genes are related. We can then identify gene regulatory interactions and map the network.

TR: What will you do with this information?

TG: We have hopes of assembling whole genome regulatory models in novel organisms, which could be very powerful. We plan to try it out on electricity-producing organisms, which produce electricity directly from carbon sources.

We will couple the regulatory network with a model of the metabolic network [a map of the cell's metabolic reactions], which is where the real business of turning carbon into electricity takes place. Then we'll try to predict what will happen if we tweak genes or nutrients. We will try to decide if and how we could increase the power output or the thermodynamic efficiency of the organism.

Understanding these networks could also help scientists build artificial circuits from scratch. Scientists have already built a number of biological machines, such as toxin detectors or bacterial cameras. That was neat circuit engineering, but most of these devices are built using just three or four component parts. Understanding gene regulators will broaden the list of parts that can be used, because scientists will understand how the parts will impact the cell.