It is indeed a rare occasion when people can get into the world’s leading scientific journal through playing computer games. But the video game Foldit is not an ordinary game. Rather, it is a software that utilizes the creativity and intelligence of tens of thousands of players in a goal of solving scientific problems.
A brainchild of Seth Cooper of University of Washington, the game was created with the goal of working out the complex 3D structure of different proteins. It is a challenging task, considering all the long chains of amino acid that are pieced together in a fiendishly complex way. To unravel it, scientists typically need to grow crystals of purified protein before bouncing X-rays off them. But Foldit takes a different approach by tapping the IQ ATM of its smart gamers. You don’t need a PhD in biochemistry to play the game. Instead of struggling with “rotating alpha-helices” and “fixing degrees of freedom”, Foldit players can simply ‘tweak’, ‘freeze’, ‘wiggle’ and ‘shake’ the on-screen shapes with the helpful aid of colorful visual cues and introductory tutorials. The game quickly became popular.
Foldit’s success relies not only on its user-friendly interface, but also on its nature as a game that both promotes competition and encourages social interaction. Players compete individually or as a team with other gamers on working out a protein structure that’s most stable. And during the process, the players can share their solutions via wiki or through online forums. By playing the game, the players get a sense of achievement and have an immersive, fun brain workout.
Foldit’s origin lies on Rosetta, a piece of software designed to piece together the long chains of amino acids by simulating and testing different combinations of structures. Rosetta relies on the principle of “distributed computing,” in which volunteers donate computing power on their computers when they don’t need it to running the program. This surplus of number crunching power helps to speed up the laborious process. But even with this, software like Rosetta, in words of Foldit player named CharlieFortsConscience, “lacks the human intuitive sense of knowing when to crunch numbers, or which tool to best use at a certain point in the process.” ( So the volunteers suggested combining their biological brain with the mechanical brain.
And they got their wish with the interactive Foldit in 2008. Foldit players incorporated a wide range of different strategies to tackle the “amino acids puzzle” with some members tackling the early stage opening, some handling the middle, and others doing the end-game polishing. And they were willing to take risk to achieve long term success. While Rosetta usually shuns a potentially interesting move if it means making the protein temporarily less stable, the live players of Foldit are willing to investigate such move because they are confident that they eventually are able to put the pieces back together in an even better state. Therefore, it is not surprising that the players outperformed Rosetta in 5 of the puzzles and matched it in 3 of them.
Cooper says the impressive result proves the hypothesis that human spatial reasoning can substantially optimize the output while retaining Rosetta algorithm as a user tool. And it is a perfect example of “cognitive surplus,” in which free time and mental energy can be channeled into something productive. Cooper thinks he can use player’s strategies to design better algorithms for later versions of Rosetta. In the meantime, Foldit players can pat themselves in the back in appearing in such prestigious scientific journal as Nature.
Works Cited
Cooper, Seth, et al. "Predicting protein structures with a multiplayer online game." Nature 466.7307 (2010): 756+.
Cooper, Seth, et al. [Supplementary Information] 1-45
No comments:
Post a Comment