Video Gamers Beat Scientists and Computers in Protein Folding Competition
Researchers from the University of Michigan (U-M) have discovered that video gamers can be as good or better than
professionally trained crystallographers and computer algorithms
at identifying the shape of a protein, a finding that has the potential
to bolster citizen science initiatives and classroom gamification.
The
researchers held a competition "to interpret biochemical data in order
to discover the protein's structure," according to a news release.
Participants included 469 gamers, two trained crystallographers, 61
University of Michigan undergraduate students with computer modeling
training, and two separate computer algorithms.
The video gamers were playing Foldit,
a computer-based videogame that challenges players to use their
puzzle-solving skills to fold 3-D graphical representations of proteins
and consequently contribute to crowdsourced scientific research into
diseases such as HIV, cancer and Alzheimer's. In the end, the video
gamers outperformed all other groups in the competition.
Analysis
of the competition results "uncovered a new family of proteins that
appears to be involved in preventing plaque formation, which is
implicated in diseases like Alzheimer's," stated the news release.
One
major difference between the winning gamers and the other groups is
that the video gamers worked collaboratively, while the
crystallographers, computer modeling students and computer algorithms
worked independently. The researchers concluded that collaboration may
improve results in this type of research.
"We think this is a
big deal because interpreting an electron-density map can be a
labor-intensive, error-prone process — and we show that crowd-sourced
Foldit players can do it as well as, or better than, professionally
trained crystallographers," said Brian Koepnick in a prepared
statement. Koepnick is a graduate student from the University of Washington Institute for Protein Design, and a member of the team that
designed the contest and analyzed its results.
Through this
competition, the researchers picked up some tips and tricks from the
gamers and plan to incorporate those findings into the software
scientists use when modeling protein structures.
Scott
Horowitz, a postdoctoral fellow at U-M and co-author of the study,
said he plans to start using Foldit in his classes. "I've seen how much players
learn about proteins from playing this game,"he told Michigan News.
"We spend weeks and weeks trying to jam this into students' brains and
Foldit players learn it naturally because it's fun."
The study
was led by U-M researchers in collaboration with the University of
Washington, University of Massachusetts-Dartmouth and Northeastern University, according to Michigan News. The study has been published in Nature Communications.
About the Author
Leila Meyer is a technology writer based in British Columbia. She can be reached at [email protected].