Algorithm discovery by protein folding game players

Abstract

Foldit is a multiplayer online game in which players collaborate and compete to create accurate protein structure models. For specific hard problems, Foldit player solutions can in some cases outperform state-of-the-art computational methods. However, very little is known about how collaborative gameplay produces these results and whether Foldit player strategies can be formalized and structured so that they can be used by computers. To determine whether high performing player strategies could be collectively codified, we augmented the Foldit gameplay mechanics with tools for players to encode their folding strategies as "recipes" and to share their recipes with other players, who are able to further modify and redistribute them. Here we describe the rapid social evolution of player-developed folding algorithms that took place in the year following the introduction of these tools. Players developed over 5,400 different recipes, both by creating new algorithms and by modifying and recombining successful recipes developed by other players. The most successful recipes rapidly spread through the Foldit player population, and two of the recipes became particularly dominant. Examination of the algorithms encoded in these two recipes revealed a striking similarity to an unpublished algorithm developed by scientists over the same period. Benchmark calculations show that the new algorithm independently discovered by scientists and by Foldit players outperforms previously published methods. Thus, online scientific game frameworks have the potential not only to solve hard scientific problems, but also to discover and formalize effective new strategies and algorithms.citizen science | crowd-sourcing | optimization | structure prediction | strategy C itizen science is an approach to leveraging natural human abilities for scientific purposes. Most such efforts involve visual tasks such as tagging images or locating image features (1-3). In contrast, Foldit is a multiplayer online scientific discovery game, in which players become highly skilled at creating accurate protein structure models through extended game play (4, 5). Foldit recruits online gamers to optimize the computed Rosetta energy using human spatial problem-solving skills. Players manipulate protein structures with a palette of interactive tools and manipulations. Through their interactive exploration Foldit players also utilize user-friendly versions of algorithms from the Rosetta structure prediction methodology (6) such as wiggle (gradient-based energy minimization) and shake (combinatorial side chain rotamer packing). The potential of gamers to solve more complex scientific problems was recently highlighted by the solution of a long-standing protein structure determination problem by Foldit players (7).One of the key strengths of game-based human problem exploration is the human ability to search over the space of possible strategies and adapt those strategies to the type of problem and stage of problem solving (5). The variability of tactics and strategies s...

Molecular simulations and visualization: introduction and overview

“…The first 'levels' of the Foldit game are designed to train the players in order to accomplish increasingly complicated tasks. Interactions among players have led to remarkable results from a biological point of view [ 106 , 107 , 108 ] but also to develop collaboratively new algorithms to solve a particular problem [ 109 ].…”

Design-driven, multi-use research agendas to enable applied synthetic biology for global health

“…Although it is not clear how many participants in the flu inhibitor game were from developing countries, this example is important because it shows that non-experts can be involved in cutting edge research. Interestingly, several of the protein folding algorithms created by the 'non-expert' gamers have been superior to those developed using state-ofthe-art computational method (Khatib et al 2011). Similarly, it is not difficult to imagine that DIY-bio communities, comprised of individuals with broad differences in levels of formal scientific training, could be readily engaged to tackle problems in global health (see Landrain et al and Betten et al in this issue for further description of DIY-bio and the role for different stakeholders in developing these technologies, respectively).…”

A GPU-accelerated immersive audio-visual framework for interaction with molecular dynamics using consumer depth sensors

“…5 In some cases, networked crowds can produce useful new strategies for molecular optimization tasks. 35 These exciting studies, at the interface of human-computer interaction, distributed computing, and molecular science, raise the prospect that distributed infrastructures along with new interface technologies can utilize the power of the internet along with crowd intelligence to solve scientic problems, simultaneously engaging the public with fundamental research questions.…”