The gamification of scientific research
Earlier this week I sat on a panel at the Open Innovation Summit at the Stevenage Bioscience Catalyst facility for a discussion around crowdsourcing in the biotechnology and pharmaceutical sectors. A central theme of my own contributions were around the rise of games that were designed to get citizens involved in the research process.
I’ve written about a few of these games on the blog before, with Phylo produced by McGill in a bid to increase the understanding of genetic research. Eyewire is a similar effort, this time produced by MIT, that aims to further understanding of neuroscience. Both have proved pretty successful, generating well over 100,000 players each in a short space of time.
Arguably the most powerful however is a game called EteRNA that aims to get citizens carrying out experiments on the folding behaviours of RNA molecules. This game is unique in that it connects gamers with an actual biochemistry lab. The initial results from the game have been put towards a paper published this week in the Proceedings of the National Academy of Sciences. What’s remarkable is that the paper had 37,000 authors, of which only 10 were professional scientists.
The game began life several years ago as Foldit, a game built out of the University of Washington to help predict the shape with which a string of amino acids will fold into. Players could tweak virtual strings, in the process collectively doing a better job than the fastest computers could at the time.
Two members of the Foldit team, Adrien Treuille and Rhiju Das, conceived of EteRNA back in 2009. “The idea was to make a version of Foldit for RNA,” says Treuille, who is now based at Carnegie Mellon University in Pittsburgh, Pennsylvania.
The new version of the game had much of the same functionality as Foldit, but this time came with a laboratory interface for designing RNA sequences for synthesis, keeping track of hypotheses for RNA folding rules, and analyzing data to revise those hypotheses.
The game design is quite ingenious. Players are asked to solve a series of puzzles involving known RNA structures. Once you’ve proved your mettle, you get the chance to join the EteRNA research team. This team works to design RNA sequences that will fold into a target structure. Eight structures are chosen each week to go off to the partner lab at Stanford University for further work, with data then returning to the game revealing accuracy and success.
The results thus far have been stunning. Over a two year period, some 4,000 players have graduated to the full research team (out of a total of 133,000). This team has generated 40 new rules for RNA folding, with each rule competing against the best computer algorithms in a series of challenge to test their validity, before then going into the algorithm themselves.
The human players trumped the computer a whopping 99% of the time, highlighting the power of the crowd in delivering scientifically useful results. The aim is to eventually automate the entire process, all the way down to a completely robotic lab.
One of the questions posed at the open innovation event this week was on the relative challenges involved in gamifying science, and it goes without saying that it isn’t an approach that will work for every kind of problem. That said however, examples such as EteRNA show what can be achieved, and there are undoubtedly many fertile areas of science that are thus far untapped.Original post