“In the past, chemists created models of molecules using plastic balls and sticks. Today, modeling is done on computers; computer simulation has become crucial for modern chemistry.” With these words, the jury for the 2013 Nobel Prize in Chemistry praised the contributions of Martin Karplus, Michael Levitt and Arieh Warshel to the vital science of molecular modeling and simulation.
The three winners’ accomplishment lies in the originality of their research method, devised more than 40 years ago. Their approach combines two levels of analysis. The first, from “classical” Newtonian physics, allows researchers to understand large molecules as a whole. The second method, quantum physics, aims to predict chemical reactions on a very small scale by simulating the behavior of atoms and electrons. While classical physics describes a molecule at rest, only the quantum approach allows scientists to understand the interactions that can take place among various molecules.
Although quantum physics is by far the most accurate, it has a major drawback: The computation time needed to replicate all of the atomic particles of all the molecules involved in the reaction is beyond the capabilities of even the most modern computers. The hybrid solution devised by the three winners overcomes this technical limitation by focusing on only some of the particles with digital, computer simulation. The results are simulations the Nobel jury described as “so realistic that they predict the outcome of traditional experiments.”
“Computer simulation allows us to observe the very structure of the protein,” said Arieh Warshel, a member of the Nobel Prize-winning team. “For example, it allows us to understand how an enzyme acts on the food digestion process. This is valuable information for future launches of new drugs.”
The scientific and human adventure that led to the Nobel Prize began in the 1970s, when Arieh Warshel joined Martin Karplus in his laboratory at Harvard University near Boston. Karplus had already developed computer-simulation programs based on the quantum approach. Warshel, meanwhile, was a recognized expert in intramolecular and intermolecular potentials (energy stored by the molecule that can be absorbed or released during a chemical reaction). And while working at the Weizmann Institute of Science in Israel, he collaborated with Michael Levitt on designing a high-performance model based on classical physics.
By 1972, Karplus and Warshel had published the first successful computer modeling of atoms, combining classical and quantum physics. One year later, Levitt joined them to work on modeling an enzymatic reaction. By 1976, their hybrid method was in widespread use worldwide.
Over the years, many improvements have enriched these pioneering efforts, taking particular advantage of the ever-improving performance of new generations of computers. “Current simulations are much less expensive and are conducted on much more powerful and complex systems,” Levitt said. “But the basic model remains the one developed in the ’70s.”
FASTER AND MORE ACCURATE
Today, computer modeling is commonly applied to all types of molecules, regardless of their size or geometry. The technique groups the atoms of a single molecule according to behavior. This enables scientists to perform quantum calculations – which consume most of the computing power – for only those parts of the molecule involved in the chemical reaction. The result saves time while improving computational accuracy.
The method has particular relevance to the pharmaceutical industry. Traditionally, developing a drug has relied on random experimentation on a wide variety of molecules, a costly and inefficient process. By 3D-modeling specific molecules and digitally simulating the mechanisms that enable or inhibit biological processes, manufacturers can better focus their research.
“I’m most proud of the fact that this method has contributed to the design of general antibody models,” Levitt said. “This research began in the ’80s and led to the creation of many of the best current anti-cancer molecules, such as trastuzumab and bevacizumab.”
Digital modeling and simulation remain as transformative today as they were 40 years ago. “I am convinced that, for the moment, simulation is the only valid approach to truly understanding biological functions at the molecular level,” Warshel said.
For the new Nobel winners, the power of computers is critical to the field’s future progress. At the same time, they stress the need to develop hybrid methods by improving quantum models.
“Chemists should understand that there are several ways to use simulation at various levels of scale,” Warshel said.
Levitt agreed. “Good models will be crucial for the evolution of chemistry, particularly on issues of ‘green’ chemistry, which are so vital to the future of humanity. Optimizing existing methods and inventing new approaches – this is a challenge that should mobilize the brainpower of a new generation of researchers.” ◆
Martin Karplus discusses winning the Nobel Prize: http://www.youtube.com/watch?v=Cq60JJ-vp2E&feature=youtu.be