set clashing to >1

Case number:845813-989614
Topic:Game: Tools
Opened by:Tony Origami
Opened on:Sunday, April 24, 2011 - 22:08
Last modified:Monday, December 31, 2012 - 23:58

Would it be easy for the foldit programmers to allow us to set clashing to a value of greater than 1? I don't know what effect that would have but it would give us more options for experimentation.

(Sun, 04/24/2011 - 22:08  |  6 comments)

Joined: 06/17/2010
Topic: General » Game: Tools

Unsure, but what you want to accomplish that way?
CI=1 means that client not allowing to create new clashes (and removing all existing one) when shake/wiggle (afik).
So higher CI will do what? Make more distances between structures?

xedr's picture
User offline. Last seen 49 weeks 4 hours ago. Offline
Joined: 05/16/2008

Wiggle tries to optimize an error function that takes into account a number of variables, including clashing. I think this would let us "mess" with the error function to good effect. if we could mess with the importance of other variables (voids, hydrophobe exposure, etc) too, I think that would also be helpful.

Joined: 12/27/2010
Groups: None

My thinking is that CI = 1 is the normal setting. If we can go from CI = 0 to CI = 2 then we are exploring from both sides of the clashing importance. What it will achieve I don't know, but I hope something good will come of it.

Joined: 06/17/2010

K, maybe s1 form dev team told us what is even possible :)
S1 told, that lowering CI is like "ignore physics".

bob1928's picture
User offline. Last seen 2 years 24 weeks ago. Offline
Joined: 12/16/2012
Groups: Gargleblasters

I think C > 1 would be a tool to help stir thing up. The gammer could probably tell it the result looks better and persure it; or go back if it looks worse.

I think 'ignoring physics' is why fold it has been so successful ... strange things get tried that would make the biochemist cringe.

The final best valid state would still be C=1.

With any respectable IDE the code should take moments to create ... at least a few days for the team to proof it (beat on it, find errors, debug, polish).

tamirh's picture
User offline. Last seen 7 years 37 weeks ago. Offline
Joined: 05/11/2012

I'll bring this up with the scientists to see if this would be useful with how Rosetta handles clashing importance values.


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, RosettaCommons