Foldit

In Short: Fold proteins. Its a fun puzzle game. You are making the world a better place (really, the real world, not a game).

Long version: Protein folding is very serious. It helps scientists understand neurodegenerative disease, conditions in which bacteria thrive or die (contrary the proteins thair made from not conforming to the same constraints), the processes of viral formation, and many other uses in medicine and industry. Trouble is, it is a very complicated problem just figuring out how many ways a protein can be folded, then on top of that, most possible folds are, for lack of a better term, quasimoto. There are very, very few combinations of folding that are useful. A long series of good folds need to be strung together to reach a valid end. A certain polypeptide chain has an estimated of 3[sup]300[/sup] possible folding formations and one known well-formed fold result. If we folded this peptide with a computer, trying a different fold every picosecond, we'd all be burned by our sun going supernova long before the computer made enough progress to measure. Nature knows the right way to get there, no single person has ever figured it out. And there are only theories as to way nature never seems to get this particular polypeptide chain folded in a wrong way. Some scientist came up with a way to get more folds done per picosecond and not waste time trying the vast majority of quasimoto folds that will never lead to a good result. The average human brain is surprisingly fairly competent at natural three-dimensional pattern matching and spatial reasoning abilities. A baby 10 seconds out of the womb gets distressed watching an Ames room; the human brain already knows when something's not 3-dimensionally prudent. So, give a 3D image of an unfolded protein to a human brain, chances are good they will be able to fold it while avoiding a fair number of invalid folds. That still leaves a really big number of folds that are not obviously easily visibly discernible as malforming. This is overcome by the fact that 240,000+ human brains are simultaneously trying folds (which are more likely to be valid than not) rather than one computer trying all the good and bad folds, one at a time. Thus we come to the Foldit game. Players are given a 3D representation of a protein and asked to find the most efficient way to fold it. They get points based on their performance. The best results are sent to other players to see if anyone can improve them. And it works great. The folding necessary to arrive at the crystal structure of the Mason-Pfizer monkey virus had stumped scientists for 15 years. They distributed all the pieces of the problem on Foldit. They gave people 21 days to make some progress that they hoped would be useful in someday finding the solution. In 15 days, players had solved the entire process. Scientists had designed an artificial enzyme to catalyse Diels-Alder reactions (widely used in synthetic chemistry). Its was terribly inefficient and lacked potency. They distributed it on Foldit. Players added to the enzyme until a combination of 13 amino acids was found to increased potency more than 18 times.

zerothis # 2014-02-18 02:19:30

hardware: x86-64 CPU,

display: textured polygons