Chemistry, Physics Nobels Awarded to Teams Supported by GPUs

by Jamie Beckett

It’s not every day your work assists someone who wins a Nobel Prize. This week GPU computing did it twice.

On Tuesday, an international team of chemists — Jacques Dubochet, Joachim Frank and Richard Henderson — won the prize for their work with cryogenic electron microscopy, which allows scientists to see the detailed protein structures that drive the inner workings of cells.

On Monday, a trio of American physicists — Rainer Weiss, Barry Barish and Kip Thorne — won science’s most prestigious honor for detecting gravitational waves, a phenomenon Albert Einstein predicted more than a century ago.

These breakthroughs are major advances in our understanding of the cosmos, and the cells inside our own bodies. They generated headlines around the world even before their work was recognized this week by the Royal Swedish Academy of Sciences, which bestows the Nobel awards.

Structure of the alcohol oxidase enzyme, visualized by Cryo-Em.
Structure of the alcohol oxidase enzyme, visualized by Cryo-Em. Image courtesy of J. Vonck, D. Parcej and D. Mills.

The World Within

The Nobel Prize for Chemistry recognizes a technique called cryo-electron microscopy that relies partly on GPUs to accelerate image processing and reconstruction of 3D macromolecular structures.

With Cryo-EM, as it’s known, researchers can now freeze molecules in mid-movement and portray them in atomic resolution to view biological processes never seen before.

This furthers understanding of life’s chemistry and is key for finding new drugs. “This method has moved biochemistry into a new era,” the Nobel Prize committee said in a statement.

In Cryo-EM, scientists capture vast amounts of high-resolution images. They use RELION (REgularised LIkelihood OptimisatioN, pronounced rely-on), a GPU-accelerated open-source software program to process and reconstruct the 3D images.

Cryo-EM is already helping scientists better understand disease. They’ve used the technique to explore the architecture of proteins that cause antibiotic resistance, to produce a 3D structure of an enzyme linked to Alzheimer’s disease, and last year, to understand the mosquito-borne Zika virus.

A research team used the same technique to figure out the structure of proteins involved with the human biological clock, an advance that was recognized with this year’s Nobel Prize in Medicine.

In 2015, the journal Nature Methods dubbed Cryo-EM “Method of the Year.”

LIGO detected the collision of two black holes for the first time.
This computer simulation shows the collision of two black holes, a powerful event detected for the first time by LIGO.  Image produced by the Simulating eXtreme Spacetimes project.

Big Picture

The Nobel Prize for Physics, by contrast, caps off the tale of a decades-long search for gravitational waves — a phenomenon predicted more than a century ago — observed for the first time with the laser interferometer gravitational-wave observatory.

The waves — ripples in the fabric of space and time — are caused by events such as colliding black holes, making the ability to detect gravity waves key to better understanding our universe.

In announcing the award in Stockholm, a Nobel Committee representative called it “a discovery that shook the world.”

In support of this effort, GPUs played a role crunching the data collected by the LIGO observatories in Hanford, Washington, and Livingston, Louisiana, making the detection of the first gravitational waves in 2015 possible.

Researchers had just started up the most advanced version of LIGO when the vibrations from a massive pair of colliding black holes slammed the detectors in Louisiana and Washington with a rising tone, or “chirp,” for a fifth of a second. The waves from that violent collision took about 1.3 billion years to reach the LIGO detectors.

In addition to confirming a core element of Einstein’s theory of relativity, the discovery pioneered a new form of astronomy based on the study of gravitational waves. Since 2015, LIGO has detected three other gravitational waves, all generated by colliding black holes.

For more information about Cryo-EM, RELION and GPUs, attend our webinar on October 18 at 11 a.m. Pacific Time.

Featured image credit: Adam Baker, via Flickr