Perhaps a future without energy terms

Case number:699969-989888
Opened by:Seagat2011
Opened on:Wednesday, June 22, 2011 - 23:16
Last modified:Saturday, June 25, 2011 - 07:30

it occurred to me while talking to anthunk regarding the role quantum entanglement plays in human "imagination" that even a mind with a powerful imagination still must be trained to follow the folding rules proteins undergo.. or atleast I thought this was true.

It may be possible to design a software which is allowed to formulate it's own rules of physics, unbounded by Rosetta energy terms, to find it's own way toward the Native fold. The software would be given no a priori knowledge of physics, but only a direct access to the NMR data (as a guide) as well as an en vivo folding environment to commence comparative folding. As the machine learns and becomes proficient, you can then ask it for it's rules. The rules it gives you will be the Rosetta terms.

(Wed, 06/22/2011 - 23:16  |  5 comments)

Joined: 12/06/2008
Groups: Contenders

"...formulate it's [sic] own rules of physics..."?

Only if you're living in some other universe. There's a reason why they're called "laws". I do believe I can hear Isaac Newton whirling like a dervish in his grave.

Dropping priority to 5-Low. Not really relevant to this game.

Joined: 12/27/2010
Groups: None

I am not sure I understand what Seagat is saying Boots but to a certain extent, tricks like setting clashing to near zero is re-writing the laws of physics.

Tlaloc's picture
User offline. Last seen 2 weeks 13 hours ago. Offline
Joined: 08/04/2008
Groups: Mojo Risin'

Sure it is. But the object is not to try to model physics...the object is to get the right answer. Reducing clash importance allows you to find a lower energy state, which is possibly closer to the native. If we could accurately model what a protein does when it folds itself, we wouldn't need foldit or rosetta, so we "cheat" and use any technique we can to find the lower energy states.

spmm's picture
User offline. Last seen 3 days 1 hour ago. Offline
Joined: 08/05/2010
Groups: Void Crushers

one of my favourite short videos is of an ''íntelligent'' robot atempting to build a column of childs blocks, something most kids learn very quickly. The robot very carefully and repeatedly puts the first block into the correct position, at the top of the pile, the block drops to the floor and rolls a short distance, the robot never learns to start at the bottom. Pehaps they do now, it is an old video. Also not sure if the monkeys have come up with any acceptable prose as yet either.
I guess what I'm trying to say is that the method above sounds very random and random usually takes a long time, doesn't mean it may not work; if you abrogate the rules of physics completly how do you know if your answer has any meaning?

Joined: 12/27/2010
Groups: None

I agree with Tlaloc - that is the point I was trying to make - that the end justifies the means. Of course Spmm is also right in that you eventually have to apply the laws of physics to the final result.


Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, Boehringer Ingelheim, RosettaCommons