Biophysics seems like a feel-good field … it’s always telling us how well-made we are. A recent piece in the Times Science section served up a crash course on that discipline, alluding to the work of William Bialek, who is a professor of physics, an architect of the Integrated Science curriculum, and apparently the happy owner of an “impish, abstractedly cerebral face and full, free-wheeling beard.”
In the article, Bialek explains why the photoreceptors in our eyes are so ideally constructed: they are designed to respond to even single photons, which are the smallest discrete units of light. “Light is quantized, and you can’t count half a photon,” he says. “This is as far as it goes.” So, at the risk of inane analogy, it’s kind of like a perfect gumball machine that would accept even pennies, accommodating the smallest extreme of currency.
That’s the basic idea behind optimization. Evolution has made some biological systems really, really, unsurpassably good at what they do, as good as the laws of physics will allow. According to the article, biophysicists have spotted such systems throughout the living world — in bacteria, in fruit fly embryos, in sharks, in us. Also,”tenets of optimization may even help explain phenomena on a larger scale, like the rubberiness of our reflexes and the basic architecture of our brain.” (Personally, I would be interested in the basic architecture of Bialek’s beard — build some sophisticated mathematical models for that puppy. You’re welcome, Biophysics Student Still Looking For A Thesis.)
Optimization analysis is cool because it allows biophysicists to 1) express deep biological principles in “an elegant set of equations,” 2) use those equations to make predictions about how real-life systems might work, and 3) test those predictions in the lab. The article shows us one example from the man himself:
In one optimization study, Dr. Bialek and his colleagues considered the dynamics of a major signaling molecule in the fruit fly embryo called bicoid.
It was known that bicoid bits were dispensed into the crown end of a fruit fly egg by the mother, that the molecules diffused tailward during development, and that the relative concentration of bicoid at any given spot helped determine the segmentation of a budding fruit fly’s form. But how, exactly, did the fly translate something as amorphous and borderless as a seeping oil spill into the ordered grid of a body plan?
The researchers calculated that, to operate optimally, each cell in the developing embryo would match the strength of its bicoid signal against an overall range of possible signal strengths, essentially by comparing notes with its neighbors. Sure enough, experiments later showed that embryonic fly cells perform precisely this sort of quantitative matching in response to a bicoid stimulus package. “It’s one of those things where we could have failed dramatically,” said Dr. Bialek, “but we succeeded better than we could have expected.”