Folding Proteins & Market Clearing
I live in an apartment off campus, which, as of the relatively cooler days over this past weekend, had yet to turn on our heating system. Since the walls are thin and my roommate and I live in the corner of the building, Saturday night got pretty cold. Sunday, however, we both slept comfortably, and the next morning my roommate touted that he had solved our heat problem: he spent the evening with an overclocked GPU from his gargantuan desktop computer resulting from him having donated his processing power to a cancer research organization which relies on massively parallel cloud computing to analyze data.
I laughed at the idea of heating our room with the only heat source which is less energy efficient than a light bulb, but it did get me really thinking about the implications of such a trade off; my roommate and this organization participated in a form of an exchange, decidedly like those which we have been studying this week in our Networks class. The idea of using the Internet to establish a mutually beneficial type of network exchange is not a new one, but the tools have evolved over time. We have learned that networks often exist to generate units of welfare, and this is no exception.
At the most basic level this is a network like the ones designed in class for the study of power in networks, but these types of transactions are much more complicated, since the consumers, like my roommate, are not locked into the exchange. If there is no need to heat our apartment, we have little incentive to raise our electric bill to destroy his processor. One organization that has managed to effectively increase incentive for potential “buyers” is the University of Washington through their creation of “Foldit”.
Foldit is a game that challenges its users to find optimal configurations of proteins so that they can achieve the lowest possible potential energy for the structure. Since even the best algorithms for solving such a problem are so complicated, effectively crowdsourcing the problem has proven to be unbelievably successful. The game was recently featured in a recent article in Nature, http://www.cs.washington.edu/homes/zoran/NSMBfoldit-2011.pdf, since its users managed, in only three short weeks, to find an optimal configuration for a protein critical to the operation of HIV, and have effectively opened the door to new possible treatments. The concept of Foldit is brilliant and represents market clearing prices at its finest: the game has become in such high demand, that it is currently maxed out on server time.
The possibilities of such an idea are endless and represent untapped markets of prospective ‘buyers’ who can provide virtually free, yet necessary services in a multitude of fields. As we have seen in Networks, however, it is first necessary to increase the worth of the ‘product’ to the Internet community in order to achieve the optimal market clearing. Until then, my roommate and I will continue to rely on our landlord to keep us comfortable.