Supercomputer helps see how humans think
Professor uses fast gadget called Frontera to explore how cells communicate in bid to cure diseases like Alzheimer’s, Parkinson’s
DALLAS — Jose Rizo-Rey wants to know how the brain gears up to think. To solve that mystery, he’s getting help from one of the fastest supercomputers in the world. It’s based in Texas.
Rizo-Rey is a professor of biophysics at the University of Texas Southwestern Medical Center. He’s interested in how brain cells send perfectly timed signals to one another so that we can ponder a particularly tricky math problem, or decide on a new job.
The problem is, those signals get sent fast — and the process Rizo-Rey wants to study happens over 1,000 times faster than the blink of an eye.
So, Rizo-Rey is using a supercomputer called Frontera to help him visualize how the brain thinks.
Frontera is “the 16th most powerful supercomputer in the world, and the fastest supercomputer on a university campus in the U.S.,” according to the Texas Advanced Computing Center in Austin.
Dan Stanzione, associate vice president for research at the University of Texas-Austin, runs the center. He said supercomputers are helpful when scientists need an answer to a complicated question quickly.
“It’s when you have a big problem, you have a lot of problems, or you have an urgent problem,” Stanzione said.
Rizo-Rey hopes scientists can apply his findings from Frontera to diseases like Alzheimer’s and Parkinson’s, where thought goes awry.
“If we understand the system,” he said, “that can always help you to design therapies for diseases that arise from defects.”
Our brain cells communicate by sending signals to one another across a small gap called a synapse. Those signals come in the form of chemicals called neurotransmitters.
Right before a signal gets sent, neurotransmitters are trapped in tiny bubbles called vesicles. These vesicles are a bit like cars at the starting line of a race. Once they get the signal to go, they merge with the edge of the brain cell, spilling their contents out into the synapse. The neurotransmitters reach the next cell, send their message, and the signal keeps on going.
“The whole system is very well regulated,” said Rizo-Rey. “And the timing is really, really fast.”
Rizo-Rey has degrees in theoretical physics and organic chemistry. He arrived at UT Southwestern from Barcelona in 1989 and established his lab in 1995.
He’s interested in what happens to vesicles right when the starting gun goes off.
Vesicles can’t “fuse” with the edge of a brain cell all on their own: They get help from a host of different proteins. One of Rizo-Rey’s colleagues, Thomas Südhof, won the Nobel Prize in 2013 for identifying some of them.
Rizo-Rey wants to know how these proteins shift and change their shape right when that fusion process happens. Trouble is, he can’t figure that out just from knowing what they look like.
Since Rizo-Rey can’t see the process for himself, he models it — using a supercomputer.
Supercomputers make calculations much faster than a desktop computer or a laptop. They’re also much bigger. Frontera, at UT-Austin’s J.J. Pickle Research Campus, is slightly longer than a large tractor trailer, and about twice as wide as one.
Rizo-Rey’s calculations on Frontera take several months to run. On a different computer, they could take years or even decades.
Watching fusion happen in real time is next to impossible, since it takes place in less than 60-millionths of a second. Instead, Rizo-Rey uses Frontera to make a model of the millions of atoms that make up a vesicle, the edge of a brain cell and the key proteins involved in fusion.
In his model, Rizo-Rey arranges the fusion proteins where he thinks they’re supposed to go. He asks Frontera how each component moves: how fast, and in what general direction.
Then, he presses play, and the supercomputer calculates what happens next.
Researchers have studied fusion using supercomputers before, but Rizo-Rey is one of the only scientists modeling it at the synapse using his highly detailed atomic model.
Harel Weinstein is a professor of physiology and biophysics at Weill Cornell Medicine who is not involved with Rizo-Rey’s work. He said Rizo-Rey’s research is a testament to the importance of using computers to understand how our brains and bodies work.
“It is certainly an extraordinarily attractive beginning to our understanding of these mechanisms of connection,” Weinstein said, “between the vesicles, and the conditions of fusion.”
Rizo-Rey’s goal is to watch fusion happen using Frontera, so he can find out how these proteins help vesicles spill their guts and keep signals going in our brains.
He said while he’s no computer expert, he’s not afraid of them, either.
“If I am 100 percent honest,” he said, “I’ve been having a lot of fun.”