Skip to main content
Premium Trial:

Request an Annual Quote

Expertise Not Required: Online Game Tackles Esoterica of Protein Structure

Premium
In the effort to mine the mysteries of unfolded proteins, researchers from the University of Washington have turned to the unlikely realm of online gaming.
 
Earlier this month, the researchers made publicly available an online game where the object is to determine the structure of unfolded proteins, launching the field of protein analysis into an area more closely associated with geeky teenagers crashing cars, killing zombies, and shooting drug dealers and pimps in the comfort of their parents’ basements.
 
Called Foldit, the game is the brainchild of David Baker, a professor of chemistry at UW and a Howard Hughes Medical Institute investigator, and David Salesin, a professor of computer science and engineering at the university. Helping to design and develop the game were Zoran Popovic, an associate professor of computer science and engineering; Seth Cooper, a doctoral student in computer science and engineering; and Adrien Treuille, a postdoc in computer science and engineering, all at UW.
 
The Foldit team also plans to roll out a protein-design game in the next few months in which players will be able to create their own proteins, some of which may eventually become the foundations for new drugs.
 
According to Salesin, the game resembles a 3D version of the game Tetris. “It’s all about fitting geometry together,” Salesin told ProteoMonitor last week. By clicking and dragging different protein parts on a computer screen, players determine the shape of protein structures. As of last week, Foldit had 18 tutorial puzzles and 63 competition puzzles. About six new competition puzzles are being released weekly. The game includes proteins whose structures are unknown as well as those that are known.
 
Baker and Salesin came up with the idea for Foldit while on a hike on Mt. Rainier in Washington state. While discussing each other’s work, the topic turned to the [email protected] project that Baker created to tap into computers and Playstations when they are not in use in order to determine the three-dimensional shapes of proteins. About 200,000 researchers are currently participating in the project, which uses Rosetta, an algorithm also developed by Baker for ab initio protein-structure prediction.
 
“I was asking David, ‘Are there situations where the computer has trouble finding things that would be obvious if people looked at it?’” Salesin said. In fact, there are.
 
While computers are very good at looking at simple proteins, as the proteins get larger their ability to determine structure becomes increasingly limited due to the growing number of mathematical possibilities. In such instances, human intuition can trump computer algorithms, and [email protected] participants were writing in saying their computers were missing things that the researchers themselves could see, Baker told Salesin that day.
 
“One of the things that was really interesting … was how into it the people who are donating their computer cycles were,” Salesin said. “They’re fiercely competitive about who’s spending more cycles on protein folding and who’s winning. And I thought, ‘What if we could actually harness that competitive energy and give them access to these [modules] and let them interact with them?’ Maybe that way we could use some of their human intuition to complement the kind of low-level geometric optimization that the computer’s able to do.” 
 

“What if we could actually harness that competitive energy and give them access to these [modules] and let them interact with them? Maybe that way we could use some of their human intuition to complement the kind of low-level geometric optimization that the computer’s able to do.”

They began devising “a kind of massive multi-player online game” that would allow players to help determine how proteins fold, Salesin said. Eventually they recruited Popovic, who at the time had been consulting Electronic Arts, which develops, markets, and distributes software games. Treuille and Cooper, who were working with Popovic, were eventually recruited as well and became the principal developers of Foldit.
 
According to Treuille, the challenge of developing Foldit was that the game had to conform to the science of protein folding. Unlike games in which a player can get attacked by a chainsaw-wielding zombie, do battle with aliens with lasers shooting out of their eyes, and get bombed by terrorists and still survive, such whimsy is not possible with Foldit.
 
“We wanted to make this fun and accessible but we had to work with the constraints of real physics,” said Treuille, who, like his fellow computer scientists working on Foldit, had only a marginal understanding of protein-folding and had to quickly become experts in the field.
 
Original iterations of the game were tested on biochemists and people “that we dragged out of the hallways in the computer-science department,” Treuille said. A few months ago, the beta test site was put online and an e-mail was sent via the [email protected] mailing list seeking interested participants to test the version.
 
Others, including people from Microsoft as well as middle-school classmates of Baker’s son, were recruited to play the beta version. Many other players became testers after discovering the game through word of mouth.
 
Foldit, which uses the Rosetta software, was designed to be simple enough so that even those who have no understanding of protein folding can play, but also challenging enough for those who study protein structures as a living. Also, rather than targeting the type of hardcore gamer who plays for days without sleeping or bathing, Foldit was developed with the casual gamer in mind.
 
“Our goal was to make a game that [would be] playable by anyone from age 5 to age 85,” Treuille said. “We very much didn’t want to target some particular sub-group with explosions or something like that.”
 
Among its 25,000 players are teenagers and university professors. The best players, Treuille said, are predominantly those with no backgrounds in biochemistry.
 
Chuck Cusack, an assistant professor of computer science at Hope College in Michigan, learned about the game at a conference last November and has since become the top-ranked player. Before he started playing he knew nothing about protein folding and even now knows “almost nothing” about it, he said.
 
“Basically the way I play the game doesn’t use a whole lot of science knowledge. That’s the way a lot of people are playing the game,” Cusack told ProteoMonitor.
 
The Foldit team is storing all the manipulations its players perform with the goal of reviewing the best ones and possibly incorporating their strategies into a computer algorithm, Salesin said. In the short time that it has been in existence — the game officially launched on May 8 — Foldit has drawn intense interest, and offers have been made to translate the game into languages other than English, said Treuille, who will start teaching at Carnegie Mellon University in the fall but will continue working on Foldit.
 
Foldit is also competing in the 8th Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction, or CASP, an international protein-folding competition, as part of Team Baker Lab.
 
The Foldit team also plans to create a protein-design game that would allow players to design proteins for new applications. “I think that actually that will work … in some ways even better [than the protein-folding game]: to sort of let people design the geometry they want and let the computer fill in the proper amino acids to realize that geometry,” Salesin said.
 
These include proteins that may not exist in nature, something that computers are unable to do, Treuille said. Such proteins could potentially form the bases for candidate therapeutics.
 
While drugs based on novel proteins and RNA sequences are being developed, doing so is extremely difficult, Treuille said, because “basically no computer algorithm can do it. So you have to rely on these experts using very esoteric software to come up with protein shapes that are stable and that [elicit the correct effect].
 
“Arguably we have 25,000 people who now have these skills to attack these problems and design novel proteins,” he said.
 
The protein-design game is expected to roll out in the fall.

The Scan

NFTs for Genome Sharing

Nature News writes that non-fungible tokens could be a way for people to profit from sharing genomic data.

Wastewater Warning System

Time magazine writes that cities and college campuses are monitoring sewage for SARS-CoV-2, an approach officials hope lasts beyond COVID-19.

Networks to Boost Surveillance

Scientific American writes that new organizations and networks aim to improve the ability of developing countries to conduct SARS-CoV-2 genomic surveillance.

Genome Biology Papers on Gastric Cancer Epimutations, BUTTERFLY, GUNC Tool

In Genome Biology this week: recurrent epigenetic mutations in gastric cancer, correction tool for unique molecular identifier-based assays, and more.