Physicist Amok

Philip at BioCurious mentions a seminar at the Chemical Biophysics Symposium in Toronto entitled “Can fundamental principles of biology be uncovered by nanotechnology.”

Biology starts out with plenty of “fundamental principles” in place. Everything it inherits from physics and chemistry can be considered a “fundamental principle.” Computer science has contributed a few things as well, but to date these tend to be idealized if not trivial.

Evolution gives us a few fundamental principles, e.g. any organism we find must be able to reproduce. That’s actually an enormous constraint. Take pug dogs. They can’t breed without human intervention. If we were trying to decipher the biology of pugs, among the first hypotheses we would generate is that there’s some external agency helping them reproduce. But these constraints don’t tell you what order genes have on a chromosome.

Marcello Magnasco once explained the point of biology to me thus: “We’re trying to reverse engineer a flying saucer.” Most beasties encode what needs to be passed on to create a similar beastie in RNA or DNA, all of them that we know of make use of protein as a structural component. However, as soon as we’re dogmatic and say, “all the alien devices have screws and bolts” someone’s going to find something welded together.

This bothers physicists a lot. I know it bothered me. I think it’s partially a result of physics training. Feynman wrote somewhere in the lectures something to the tune of: physicists take the simplest possible manifestation of a phenomenon and call that physics, then shove the rest off into engineering.

In biology we start from the engineering and work back to the physical laws. Systems biologists have lost sight of this. The whole field is predicated on a desparate hope that everything you might ever want to know about a biological system is encoded in a directed graph a geneticist put together. The complete failure of this program leads to absurdities like Michael Elowitz’s continuing study on noise. I mean no insult to Elowitz. He’s the best of the bunch, and he’s played the current fad to get a professorship at Caltech. But if he were being judged on how probing a question he was asking to reduce the problem to the physics underneath, I think it would seem pretty boring.

So can we please stop seeking for fundamental principles of toaster oven design?