Monday, August 09, 2004

We have the technology, we can build it

... or, rather, "we're working on the technology, so we can build it". "It", in this case refers to a biological system that does, well, anything we'd like it to do -- churn out drug precursors more cheaply than existing chemical synthesis processes, kill a cell that's divided more than a certain number of times [like a cancer cell], chew up plaque building up on artery walls etc.

That's the kind of stuff that I had a chance to talk about today with Drew Endy, one of the folks at MIT working on what's being called "Synthetic Biology" and whose lab I'm thinking about working in. The basic idea is pretty straightforward: wouldn't it be nice if we could build a system out of biological parts [ie cells, biomolecules and their reaction networks] in the same way that we can currently build electronic systems out of silicon chips ? The implementation of that idea, on the other hand, isn't -- we don't understand biological systems nearly as well as we understand the physics of transistors, there's lots of cross-talk between the chemical reactions going on in a cell, we don't know all the reactions going on in the cell, or even what molecules are participating in them, biological systems are probabilistic ... the list goes on. And, the kicker, something that you don't have to deal with at all in electronics or software is -- even if you build a system that actually works in a cell, it evolves. Somewhere along the way, during cell division, a mutation is going to occur that totally wipes out your carefully engineered mechanism, whether you like it or not. How do you build a system that can deal with random mutations like that ?

The current state of the art is basically people building one-off systems, with results that may or may not be transferable to other systems. What Drew Endy and other people like Tom Knight [somebody else whose lab I'm interested in working in] are working on is a "platform" for engineering biological systems, with interchangeable parts with known input/output characteristics that can be combined to build novel systems out of these "off-the-shelf" parts.

One of the things that I always liked about programming was "building stuff that does stuff", and so the notion of tinkering with such fundamental, very malleable blocks and building a broad platform for this sort of biology-based engineering is very exciting to me. Luckily my program gives me the chance to do 3 short "rotations" in different labs before I pick an advisor, so I'll have a chance to try this on for size before actually committing a few years of my life to it.

[For anybody interested in reading more about this, Science News, the EE Times, the Lawrence Berkeley labs and Scientific American all have stories on this.]


Post a Comment

<< Home