A fine day for molecular dynamics

Generally speaking, I am not a “morning person.”  Like most graduate students, I tend to stay up till 1–2 AM and wake up between 8–9 AM.  But this morning, at 6 AM, my phone buzzed with a text.  A friend and co-worker of mine in India (where I suppose it was a ‘sane’ time) messaged me:

BTW, Warshel got the NP!

Normally, my reaction to my phone buzzing early in the morning is that I hazily read it, go back to sleep, and read it again when I’m in my right mind.  But this time, I was abuzz, and I was pretty sure going back to sleep was not going to happen.  Is it slightly embarrassing that I was so excited that Arieh Warshel, computational enzymologist, just won the Nobel Prize?

So here’s the back story.

My research in graduate school has focused on trying to understand the physical origins of enzymes’ catalytic power.  Enzymes are Nature’s miracle catalysts, allowing chemical reactions that would otherwise take longer than the age of the known Universe to take place in split seconds.  No man-made catalyst does nearly as well, and despite the incredible break-throughs in biochemistry, the active sites of enzymes remain as secret and mysterious as dark caves (which is actually what they normally look like in X-ray structures).

When I first started graduate school, I decided that I needed to switch my focus from synthetic catalysts to enzymes, and one of the first papers that I read on the subject that got me excited was:

“Electrostatic Basis for Enzyme Catalysis.” Ariel Warshel, et al. Chem. Rev.2006, 106, 3210–3235.

What the paper claimed is that the electrical forces that enzyme active sites exert onto their bound substrates are responsible for driving the substrates to react quickly and efficiently.  The authors used powerful computers in order to model the active sites of many enzymes and found this is a general aspect of enzyme catalysis.  In graduate school, my research has focused on testing these claims experimentally, which has so far only been supported with computer models.  Nevertheless, Warshel’s work played a significant role in shaping my research interests and graduate school project – so you can see why this day has been an exciting one!

More broadly, the three scientists who will share the prize are known as founders of a whole constellation of computational approaches that I and many other chemists use in our research, known as molecular dynamics – or MD for short.  So what’s all this MD-business about anyway?

Molecules come in many shapes in sizes – the simplest one is H-H (a molecule made out of two hydrogen atoms with a chemical bond between them), and enzymes are among the more complicated ones (consisting of tens of thousands of atoms, mostly carbon, hydrogen, oxygen, and nitrogen).  In general, you need quantum mechanics to explain just about everything about molecules.  In fact, using the classical physics we learn in high school, you would predict that atoms could never come together to form molecules (and physicists doubted the existence of molecules until the early 20th century).  Quantum mechanics is powerful and highly experimentally-validated: it makes many accurate predictions about molecules, including the way they move and interact with each other.  But quantum mechanics comes with a price: it’s very complicated – which means it takes a really long time to calculate the full quantum mechanical answer to a chemical problem.  Even the world’s most powerful computers can scarcely calculate quantum mechanical solutions for molecules with much more than 100 atoms.  That’s a far cry from an enzyme, with 1000’s of atoms.

So what can we do?  The answer is that we cheat.

If you were to perform a bunch of quantum mechanical calculations on H-H, you would find two important discoveries: first, that H-H is most stable when the atoms are ca. 1 angstrom apart (an angstrom is 10^-10 meters) – this is called the preferred bond length.  As you push the H-atoms apart or squeeze them in, the energy goes up as the square of the distance that you perturbed them away from 1 angstrom.  In other words, H-H is exactly like a spring.  In classical mechanics, we learned about springs and how they can be described by Newton’s equations of motion along with Hooke’s law.  So what this means is that we can table the whole quantum mechanics business and just treat H-H as if it were a spring, a much easier thing to think about.  Moreover, a computer would take a fraction of a second to calculate a spring energy, whereas it might take days to do the full quantum mechanics problem.  We now have a convenient short-cut!

Bigger molecules like enzymes have lots of atoms with many bonds, but we can take this “spring idea” and just keep on building on it, turning all the bonds into springs.  This way of turning a quantum mechanical problem into a classical problem is known as molecular mechanics, and the specific parameters that are needed to get this quantum-to-classical mapping right (without losing too much in translation) is called  a force field.  Don’t be confused! – this isn’t the same thing as a force field in Star Wars… although they are easily just as exciting and it would certainly be fair to say that the three Nobelists announced today – Michael Levitt, Arieh Warshel, and Martin Karplus – are chemistry’s equivalents of Jedi masters ^_^.  Jedis and springs aside, this is an exciting time for molecular dynamics.  We are at the point of figuring out how Nature’s complicated molecular machines work at the physical level, and this is helping us design new molecules for useful purposes (like medicines and fuels) using our brains instead of just by guesswork.

One comment

Leave a comment