What is actually happening at the atomic scale when two elements react? This year's chemistry prize goes to three theoretical chemists who devised a way for computers to model and predict how such reactions take place: Martin Karplus of Harvard University and the Universite de Strasbourg in France; Michael Levitt of the Stanford University School of Medicine and Arieh Warshel of the University of Southern California. Or, as the Nobel Committee put it in awarding the prize: "for the development of multiscale models for complex chemical systems."
The key was physics—specifically, finding a way to use both a quantum mechanical understanding of individual atoms in the most critical area of a molecule (which requires a great deal of computational power) and then simpler, easier-to-calculate classical mechanics to deliver the rest of the system. That combination enables a computer to model in great detail, say, the catalyst in a particular enzyme as the electrons leap from the orbit of one nucleus to another during the formation of chemical bonds , while allowing simpler calculations for the rest of the complex molecule.
"It's like seeing the watch and wondering how it works," said Warshel via telephone during the press conference presenting the prize. "What we developed is a way that requires a computer to take the structure of a protein and then to eventually understand how exactly it does what it does." Such an understanding of how things work, for example, can then be used to "design drugs or, like in my case, to satisfy curiosity." In other words, chemists no longer only experiment in the lab; they can also experiment in cyberspace.
Such theoretical modeling is already in wide employ in the pharmaceutical industry, helping to deliver drugs and is also being used to help unlock the secrets of photosynthesis—the chemical transformation of carbon dioxide to carbohydrates using only the energy provided by sunlight. A better understanding of that chemical reaction, and the molecules involved—which, after all makes the majority of life on Earth possible—could help deliver clean and abundant energy for human purposes. And Levitt, for one, would like to one day simulate an entire living organism at the molecular level.
From our Scientific American Digital archive:
Simulating Water and the Molecules of Life by Mark Gerstein and Michael Levitt, November 1998
If you have an institutional subscription, you can access our deep archive:
The Dynamics of Proteins by Martin Karplus and J. Andrew McCammon, April 1986