Placeholder image of a protein
Icon representing a puzzle

618: Bonus Impossibly Difficult CASP10 Puzzle

Closed since about 10 years ago

Advanced Overall Prediction Symmetry


August 15, 2012
Max points

The last CASP10 puzzle just closed, so we thought we'd post this tough target (for anyone going through CASP10 withdrawal) that we weren't able to give you during CASP. This refinement target, TR699, has a starting GDT_TS=85 and "should be treated as a dimer". Oh yes, and it's 234 residues! Read the puzzle comments for more details and how to deal with large proteins.

Top groups

  1. Avatar for Anthropic Dreams 100 pts. 22,380
  2. Avatar for Contenders 2. Contenders 85 pts. 22,168
  3. Avatar for Beta Folders 3. Beta Folders 72 pts. 21,966
  4. Avatar for foldeRNA 4. foldeRNA 61 pts. 21,943
  5. Avatar for Go Science 5. Go Science 51 pts. 21,821
  6. Avatar for Void Crushers 6. Void Crushers 42 pts. 21,702
  7. Avatar for L'Alliance Francophone 7. L'Alliance Francophone 35 pts. 21,540
  8. Avatar for Hun-Magyar Csapat 8. Hun-Magyar Csapat 28 pts. 21,522
  9. Avatar for FoldIt@Netherlands 9. FoldIt@Netherlands 23 pts. 21,277
  10. Avatar for Brony@Home 10. Brony@Home 19 pts. 21,197

  1. Avatar for harp
    1. harp Lv 1
    100 pts. 22,295
  2. Avatar for BitSpawn 2. BitSpawn Lv 1 99 pts. 22,150
  3. Avatar for Hanto_FZ140E_4G 3. Hanto_FZ140E_4G Lv 1 97 pts. 22,039
  4. Avatar for gmn 4. gmn Lv 1 96 pts. 22,023
  5. Avatar for stg123123 5. stg123123 Lv 1 94 pts. 21,966
  6. Avatar for bertro 6. bertro Lv 1 93 pts. 21,943
  7. Avatar for gitwut 7. gitwut Lv 1 91 pts. 21,901
  8. Avatar for mimi 8. mimi Lv 1 90 pts. 21,873
  9. Avatar for doug.ehlert 9. doug.ehlert Lv 1 88 pts. 21,848
  10. Avatar for Susume 10. Susume Lv 1 87 pts. 21,790


beta_helix Staff Lv 1

In CASP9 we had to skip these types of targets, because symmetry was not possible with Foldit… but that is no longer the case!

We wanted to give you the opportunity to fold TR699 in the context of a dimer, but we realize that this was already a big protein, and symmetry makes it even bigger.

This is why it is only worth 70 global points, think of it as an "extra credit" puzzle… it's just for fun so don't be too frustrated if it barely runs on your computer.

Some possible hints

I recommend turning off all sidechains, using Cartoon Thin, with Hydro Color and only showing Clashes. (If this is still too slow, you can try using "Line")

Since it forms a dimer, you probably want to look for a hydrophobic patch where the two sections come together. Coloring by Hydro can help you try to get as much orange at the interface.

If you have trouble loading this puzzle, try checking the box "disable network timeouts" when you login.

Please try to have fun with this one.
If it frustrates you, then at least be happy we didn't post it during CASP10!
This is the page for this CASP10 refinement target:

Note that the organizers gave us this hint:

"Starting model GDT_TS=85. Target should be treated as a dimer. "

GDT_TS=85 essentially means that ONLY 85% of the model superimposes correctly onto the native structure:

Here is the sequence logo predicted by the SAM server:

H = helix
E = sheet
C = loop (or coil)

The taller the letter at each position, the higher the probability of that specific secondary structure for that amino acid.

For example, the Valine at residue 44 is highly predicted to be part of a helix. However, the Valine at residue 56 is predicted to be in a sheet.