“Scientists Are Designing Artisanal Proteins for Your Body,” writes Carl Zimmer in the New York Times, reporting that “researchers have learned to create custom versions not found in nature.”
The complexity of interatomic forces, says Zimmer, makes it a “staggering molecular puzzle” to predict how (or if!) a protein sequence will fold up into a protein and function. From my background in molecular physics, I can second that: it’s hard to communicate just how finicky a puzzle it is to calculate these forces (Carl mentions only a few), especially those that involve multiple atoms, and those that straddle the boundary between different types of electronic physics. Then there is the vast combinatorics of the different ways proteins could arrange themeselves. Together these make the “folding problem” an extremely difficult one.
But we are making great progress. After many thousands of man-hours of research, and zillions of CPU-hours on borrowed computers, biochemist David Baker of the University of Washington claims we have basically nailed it.
“There are subtleties going on in naturally occurring proteins that we still don’t understand,” Dr. Baker said. “But we’ve mostly solved the folding problem.”
He thinks that natural proteins are not designed, and so we should be able to do better:
“There’s a lot of things that nature has come up with just by randomly bumbling around,” he said. “As we understand more and more of the basic principles, we ought to be able to do far better.”
Don’t be fooled. If it’s difficult for us, it is great big wall for random bumbly evolution. Doug Axe and others have written about that. The truth is, if it takes a lot of design effort now, it probably took a lot of design effort before. Moreoever, Dr. Baker acknowledges that the kind of proteins we can make are much shorter than many that exist in nature, and we aren’t really at the stage of making molecular machines, so it’s not yet clear at all whether we will be able to do better than those primordial designs. It is possible, maybe if we focus on different goals and design constraints. But there isn’t much reason yet to suppose that existing biomolecules can be bested. We’ll see.
It is interesting, though, that everyone agrees intelligent design is more powerful than natural evolution. Why is that? Both processes explore the same space of possibilities. Intelligent design can see further ahead, but evolution has much more time, and many more chances to win.
The big difference is that although naturalistic evolution can search for solutions, it doesn’t learn how to search. Over at the Evolutionary Informatics Lab, you will find a number of papers explaining why successful (defined as better-than-random) search for solutions requires Active Information, which embodies applied knowledge about the problem that needs to be solved.
Over the last few decades, Dr. Baker and his fellow scientists have not merely been mutating proteins randomly, but, rather, looking for insights and design principles. They increase our understanding of protein design all the time, and feed this back into refining the algorithms and design procedures. Natural selection doesn’t do anything like that. Instead, intelligent agents have been accumulating Active Information, just like the theory of intelligent design predicts.