1028: Revisiting Electron Density Puzzle 730

Closed since about 11 years ago

Summary

• Created: December 20, 2014
• Expires: January 04, 2015 at 14:00 UTC
• Max points: 100

Description

We're revisiting this old electron density puzzle to test some recent changes to Foldit! In this puzzle, there are bits of density which don't belong to the copy of the protein that you're working on. You'll have to decide which bits of density to match with, and which bits to ignore. Try to fit the protein into the density to get the best score! Secondary structure predictions are posted in the puzzle comments along with info about available tools in the game.

Top groups

• 1. Void Crushers 100 pts. 12,690
• 2. Anthropic Dreams 83 pts. 12,630
• 3. Gargleblasters 69 pts. 12,548
• 4. Beta Folders 56 pts. 11,655
• 5. Contenders 45 pts. 11,505
• 6. SETI.Germany 36 pts. 11,495
• 7. Go Science 29 pts. 11,449
• 8. L'Alliance Francophone 23 pts. 11,365
• 9. Deleted group pts. 10,952
• 10. Deleted group pts. 10,864

Top soloists

• 1. spmm 100 pts. 12,690
• 2. BitSpawn 98 pts. 12,630
• 3. frood66 96 pts. 12,434
• 4. Susume 94 pts. 12,216
• 5. Timo van der Laan 92 pts. 12,183
• 6. retiredmichael 90 pts. 11,649
• 7. grogar7 88 pts. 11,575
• 8. mimi 86 pts. 11,499
• 9. Bautho 85 pts. 11,495
• 10. Galaxie 83 pts. 11,436

Comments

[bkoep]

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.

The above predictions are already on the starting model (based on the horizontal lines in the above figure).
______________

Cutaway view hint:
If you hover your mouse over a segment and type shift-Q, the game will hide parts of the protein and cloud that are not in the same plane as that segment - it cuts down on how much cloud you see at one time and makes it easier to focus. Q returns to normal view.

[Timo van der Laan]

This is the third repeat of an ED puzzle. I would really like to see one that hasnt been solved yet.
BTW, these kind of puzzles are now very good playable, congrats to the devs on the speed improvement since NC.

[bkoep]

The difficulty is finding test cases appropriate for Foldit. Most unsolved cases involve low-resolution density maps for very large proteins.

We've been running these revisited puzzles to test new ED code and recalibrate density scoring, but we're also anxious for the next unsolved density map!