puzzle picture
505: Top CASP ROLL Puzzle 2 Predictions
Status: Closed


Name: 505: Top CASP ROLL Puzzle 2 Predictions
Status: Closed
Created: 01/20/2012
Points: 100
Expired: 01/26/2012 - 01:00
Difficulty: Advanced
Description: We want to let you work on the top 5 Foldit predictions for the second CASP ROLL target (the 5 models we would submit to CASP). Instead of giving you those final 5 models, however, we are providing you with early states of those models so that you don't start trapped in local energy minima. Resetting the puzzle will cycle through 5 player predictions. They are also available in the Alignment Tool so you can use partial threading. This puzzle will have 2 rounds, more details are in the puzzle comments.
Categories: Overall, Prediction

Top Groups

2Void Crushers12,66185
4Anthropic Dreams12,64060
5Go Science12,63650

Top Evolvers

Top Soloists

Need this puzzle? Log in to download.  


beta_helix's picture
User offline. Last seen 14 hours 21 min ago. Offline
Joined: 05/09/2008
Groups: None
Details and motivation for this

First the details for this puzzle:

This puzzle will only be up for 5 days (since the CASP deadline for this target is Feb 1st) at which point we will repost this exact puzzle but include the 5 Rosetta Server predictions. You will be able to load in any predictions from this puzzle into that puzzle as well (so you can use partial threading).
That puzzle will be up until January 31st.

Motivation for this puzzle:

Previously in Foldit we tried posting All-Hands rounds, where everyone could work on the top scoring Foldit predictions from a previous puzzle.
Unfortunately these top scoring models had already been heavily minimized that they became very difficult to improve on, and attempts to change the structure resulted in much lower scores.

This was also a problem in CASP9 with the de-novo and refinement targets that started from Rosetta Server Models. This is mentioned in the NSMB paper: http://depts.washington.edu/bakerpg/drupal/system/files/NSMBfoldit.pdf and in that paper we discussed how perturbing starting models helped in the last refinement target, TR624, so we wanted to do that here as well.
Unlike refinement, we have no idea which sections are correct or incorrect so we can't decide which regions to perturb. Foldit players have suggested posting early versions of their final top scoring solutions so that is what we are trying here. We took the top 5 scoring predictions from Puzzle 497: CASP ROLL Puzzle 2b and are giving you earlier states of those models.

This is harder to do with the Rosetta Server models, so we're going to give those to you last and hopefully you can use partial threading with your own solutions to improve on Rosetta.

Additional information on partial threading here:


Please make sure to try out all 5 starting server models, since we don't know which one is closer to the correct native structure.

We have also included an extended chain in the Alignment Tool for anyone who needs it.

User login
Download links:
  Windows    OSX    Linux  
(10.12 or later)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Social Media

Only search fold.it
Other Games: Mozak
Recommend Foldit
Top New Users

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