quicker way to copy secondary structures on freestyle puzzle

Case number:845813-987504
Topic:Game: Tools
Opened by:Perzik
Opened on:Friday, March 26, 2010 - 07:13
Last modified:Tuesday, August 23, 2011 - 20:16

Isn't there a way in freestyle puzzles (where you get the secondary structure predictions in the comment of a post) to quickly apply these structures on your protein. What I now do is look at the prediction, e.g. third amino acid is helical, go to puzzle and manually change to helical,...a bit tedious..
Couldn't we use something like the align tool where you then can thread your protein against the secondary structure prediction (instead of a known sequence and 3D structure). Or just be able to see the sequence like in the align tool and be able to click on the sequence and say which secondary structure is must have. Hope this is all clear what I mean. I'm just looking for a way out to manually assigning sec structures on a freestyle puzzle..

(Fri, 03/26/2010 - 07:13  |  4 comments)

Madde's picture
User offline. Last seen 1 year 38 weeks ago. Offline
Joined: 05/29/2008
Groups: Void Crushers

Normally the community is very fast in programming a corresponding script.
For puzzle 278 you may use lilovip's recipe: http://fold.it/portal/recipe/2264
(Keep in mind that these predictions are not carved in stone.)

beta_helix's picture
User offline. Last seen 7 hours 35 min ago. Offline
Joined: 05/09/2008
Groups: None
Status: Open » Open

Thanks for that quick reply Madde!

Indeed that is our fear with creating a magic button that will assign the secondary structure since these are predictions.

Just like the starting Rosetta predictions that we give you as starting puzzles, these secondary structure predictions are just starting points for you to try.

If a long stretch is highly predicted to be a helix, then it probably is (but it could still be wrong!), but what about regions that are half-predicted to be sheet and half-predicted to be loop?

Hopefully you have seen with the Quest to the Natives (the ones that had freestyle puzzles before them) that the secondary structure predictions are just that: educated guesses!

Perzik's picture
User offline. Last seen 3 years 3 weeks ago. Offline
Joined: 05/13/2008
Groups: Boinc.be

true that they are only predictions, but using homologous sequences as templates is also only a guess that you take.. the tool could be made such that you see the prediction and choose to either believe it for that particular region of the protein, or not..

but it's true that the community does a nice job making scripts in that regard, so it's possible to circumvent the whole manual structure assignment..

Joined: 06/17/2010
Status: Open » Closed

Predicted structure is assigned to puzzle now. Done.


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