One more book to pile on my to-read list. Via Carl Zimmer (go follow the link for a bloggingheads video interview), a fascinating book on bioengineering, Learning to Fly, by Rob Carlson, is coming out this fall.

He has some insightful thoughts:

Explicit “hands on” molecular manipulation of genomes began only in the mid-1970s, and we are still learning the ropes. Most genetically modified systems do not yet work entirely as planned. Biological engineering as practiced today proceeds by fits and starts, and most products on the market today result from a process that remains dominated by trial and error. The primary reason that the engineering of new organisms has been slow in coming is that simply understanding naturally occurring organisms remains hard.

As Carlson notes, much of the major progress has come from our increasingly facile ability to manipulate the physical stuff of life - we can clone and synthesize genes, and put those genes into all sorts of different organisms. The cost of those tools is dropping rapidly, so Carlson says the biohackers are on their way:

Stewart Brand, founder of the Whole Earth Catalog, organizer of the first Hackers’ Conference in 1984, and co-founder of The WELL and the Global Business Network, wonders: “Where are all the green biotech hackers?”[4] To which I answer: They are coming. The tools necessary to understand existing systems and build new ones are improving rapidly. As I will discuss in Chapter 6, the costs of reading and writing new genes and genomes are falling by a factor of two every eighteen to twenty-four months, and productivity in reading and writing is independently doubling at a similar rate. We are just now emerging from the “slow” part of the curves, by which I mean that the cost and productivity of these technologies are now enabling enormous discovery and innovation. Consequently, access to technology is also accelerating. “Garage biology” is here already;

Garage biology is here already - it happens in molecular biology labs all over the world, and maybe soon hobbyists will soon do this stuff in real garages.

The problem is this: the professionals are still doing just garage biology. Bioengineers make great life hackers - they can tweak existing systems, and with enough trial and error, they can achieve some amazing results.

What we don't have are the Boeing engineers of biology - the pros who can do things that you can never accomplish in a garage. My problem isn't that I'm impatient - it's that I don't see how to get there from here. I don't see how we'll get from Edison-like tinkering to computer-aided-design of complex biological systems. Can we reprogram a developmental system to do something completely new - create a new body plan? No, not now. Can we design genes or metabolic pathways with completely novel functions, from scratch (not just by small modifications of an existing gene or pathway), or create novel protein domains? Very rarely, with a lot of luck.

To do these things, we need some sort of bioengineering mathematics, some sort of formalism. For the past several decades, the key challenge has been to develop methods to manipulate the molecules of life, and to make those methods affordable. Clever bioengineers are meeting that challenge. The next step, however, looks even more daunting: finding a tractable, practical theoretical framework on which we can rest future biotechnology.

Some further reading: Here's one approach to developing a modeling framework for describing and reasoning about biological systems: Rule-Based Moedlling of Cellular Signalling, Vincent Danos, Jerome Fere, Walter Fontana, Russell Harmer, and Joan Krivine.