The Rosetta Protein Folding Game

Pumpkinstew

Pumpkinstew

Pumpkinstew

Soldato
Joined
19 Jan 2005
Posts
2,729
Location
Oxford/Bristol
Now you have the chance to prove you're smarter than your Quad.:D

The Rosetta at Home team have designed this pretty neat game where you nudge and tweak a protein molecule into it's tightly packed lowest energy state and earn points for doing so.
It's a 50meg download and it looks like there are only Windows and Mac apps for now. Get it here. http://fold.it/portal/adobe_main/

Have a bash at the tutorial first before jumping straight onto an active molecule. It helps. (Believe me :o)

Pretty soon molecules will be released for the CASP08 challenge giving users the chance to prove they can beat the brute force mass computing approach.

If enough people join we can form an OcUK group and fold as a team.

My best score on 51 so far is 9547.

(for some reason my first attempt to log in always errors but the second is OK:confused:)

EDIT: Lot's of good info here about the science behind FoldIt! http://fold.it/portal/info/science
 
Last edited:
^^
I thought that but it's a doddle. You just nudge stuff around press a button and see if the score goes up.

8969 on the second one. 102nd position. :)
 
I'll try this when i get back from the bank holiday weekend, but i'd definately be up for this. I've been playing a lot of flash games recently anyway and this sounds like an interesting challenge :-) especially if it helps the folding cause.
 
The point, as it were, is whether us mere mortals can do a better job with instinct and intuition than several petaflops worth of computing power. The problem is that the computing solution uses brute force - i.e. try every possibility and see if it works.

Consider Chess - how come one mere mortal (albeit very clever mortal) can beat a huge supercomputer. Same problem - the computer sees only numbers, whereas the human has years of practice, and intuition as to which is likely to be the best move.
 
Cob said:
I fail to see the point in that tbh. Surely F@H has that one covered.
Click to expand...

The link escapes me at the moment because I have been watching the Championship play off and the Heineken cup final and some Munstermen swarmed by table and insisted on supplying me with Kronenbourg for the intrusion... but there is a difference between what F@h and Rosetta and Predictor are doing although all three relate to protein folding.
 
Berserker said:
The point, as it were, is whether us mere mortals can do a better job with instinct and intuition than several petaflops worth of computing power. The problem is that the computing solution uses brute force - i.e. try every possibility and see if it works.

Consider Chess - how come one mere mortal (albeit very clever mortal) can beat a huge supercomputer. Same problem - the computer sees only numbers, whereas the human has years of practice, and intuition as to which is likely to be the best move.
Click to expand...

Fair point, but the computer wins at chess now.
 
Berserker said:
The point, as it were, is whether us mere mortals can do a better job with instinct and intuition than several petaflops worth of computing power. The problem is that the computing solution uses brute force - i.e. try every possibility and see if it works.

Consider Chess - how come one mere mortal (albeit very clever mortal) can beat a huge supercomputer. Same problem - the computer sees only numbers, whereas the human has years of practice, and intuition as to which is likely to be the best move.
Click to expand...

I really have no idea what you're on about.


But I'm putting it down to being an Irishman on Heinekin Cup final day.


*edit*


Aaah the game :o
 
Last edited:
Pumpkinstew said:
The link escapes me at the moment because I have been watching the Championship play off and the Heineken cup final and some Munstermen swarmed by table and insisted on supplying me with Kronenbourg for the intrusion... but there is a difference between what F@h and Rosetta and Predictor are doing although all three relate to protein folding.
Click to expand...

Ah. I'll have to have a look into it.
 
Last edited:
hmmmm, surely the computer going to wing at folding, the time you spend moving the protein round, then correcting the errors, then moving again, then finding something else wrong..... in that time the cpu have done twice as many variations surely....
 
It's a very interesting experiment for sure - plus the fact it uses mainly human brain-time rather than CPU cycles means there's no real problem running both :)


May give it a try tomorrow when back on the big screen :p


edit: hmmm well it's kinda interesting but not sure I've really taken to it, plus as it's flash-based it was quite insistent on using pretty much half one of the cores of my quad so would have a fair effect on anything using CPU time also :o
 
Last edited:
Its actually better doing it yourself cuz you can learn from your mistakes.

Results have been published recently in the latest Nature journal

Notice that if you have played the game then you have your name on the paper as last author ;)
 
You must log in or register to reply here.
Back
Top Bottom