1378b: Hemagglutinin Binder Design: Loop Graft

Closed since almost 9 years ago

Summary

• Created: May 16, 2017
• Expires: May 22, 2017 at 23:00 UTC
• Max points: 100

Description

Note: This puzzle replaces Puzzle 1378, which was incorrectly calibrated for scoring. Players may load in solutions from Puzzle 1378.



Design a protein that contains the provided hemagglutinin-binding loop! Hemagglutinin is a protein from the influenza virus. It is displayed on the surface of the virus capsid, and allows the virus to recognize and infiltrate human cells. One particular region of the hemagglutinin protein undergoes a dramatic shift during infection; it's thought that stabilizing this region (e.g. with a protein binder) can prevent the normal function of hemagglutinin and stop infection. In fact, during an influenza infection, the immune system can sometimes raise antibody proteins that bind this portion of hemagglutinin to defend against further infection. In this puzzle, we've provided a portion of the hemagglutinin protein, as well as a hemagglutinin-binding loop that is derived from one such antibody protein. We're asking Foldit players to design a new protein that incorporates this same loop!



The backbone and certain residues of the loop are fixed, and the loop is held in place with strong constraints. The starting structure of the design has 53 residues, but players may add residues up to a total of 63. Players may also delete and redistribute residues on either side of the prescribed binding loop. See the puzzle comments for filter details.

Top groups

• 1. Anthropic Dreams 100 pts. 11,113
• 2. Contenders 76 pts. 10,649
• 3. Beta Folders 56 pts. 10,598
• 4. Go Science 41 pts. 10,583
• 5. Void Crushers 29 pts. 10,480
• 6. HMT heritage 20 pts. 10,388
• 7. Deleted group pts. 10,104
• 8. Gargleblasters 9 pts. 10,097
• 9. L'Alliance Francophone 6 pts. 10,065
• 10. Natural Abilities 4 pts. 9,944

Top soloists

• 1. markm457 100 pts. 11,103
• 2. eusair 97 pts. 10,867
• 3. johnmitch 94 pts. 10,804
• 4. tokens 91 pts. 10,694
• 5. Bletchley Park 88 pts. 10,572
• 6. alcor29 85 pts. 10,568
• 7. hpaege 82 pts. 10,546
• 8. LociOiling 79 pts. 10,519
• 9. Timo van der Laan 77 pts. 10,480
• 10. retiredmichael 74 pts. 10,470

Comments

[tokens]

The puzzle comments say: "Players may also delete and redistribute residues on either side of the prescribed binding loop."

This is not possible for me. I could add residues, but not delete any residues, not even after I had added some.

[bkoep]

That's unintended behavior. Can you delete specifically those residues that you added?

[tokens]

Ahh, yes, I can delete the residues that I added, but not any other residues.

[bertro]

see:
http://fold.it/portal/node/2003808

[LociOiling]

Both rebuild and remix recipes seem to hang on the first rebuild or remix, never advancing. Cancel does not work, the message "cancelling" appears in the upper left, but it hangs there indefinitely.

This problem can be produced a manual remix. After selecting segments and starting remix, the first remix result is displayed, but remix hangs when you try to advance to the next result. Clicking the "X" to close remix has no effect. The only option is to close the client window, which seems to work normally.

In log.txt, a manual remix shows these results:

***** STARTING THREAD ActionBuild
lo,hi,size: 395,399,6
19 fragments found for fragment size 6
protocols.loops.loop_closure.ccd.RamaCheck: Initializing starting_rama_scores with current conformation.
SRVR_THRD getting notifications...
Sending SOPs:

Sending SOPs:

Couldn't quit: Illegal operation for this state (..\..\..\source\src\game\util\IRC.cc:375)
closing boinc...
Sending SOPs:

Sending SOPs:

exiting...
done.

The "illegal operation" message appears at the point where you manually close the client.

Attempting a manual rebuild crashed the client with the following in log.txt

***** STARTING THREAD ActionRebuild
protocols.loops.loop_closure.ccd.RamaCheck: Initializing starting_rama_scores with current conformation.

UNHANDLED EXCEPTION
  1: library_main +37712549 bytes (no line)
  2: library_main +28085694 bytes (no line)
  3: library_main +28085427 bytes (no line)
  4: library_main +28084834 bytes (no line)
  5: library_main +21461506 bytes (no line)
  6: library_main +21462889 bytes (no line)
  7: library_main +19008449 bytes (no line)
  8: library_main +17555767 bytes (no line)
  9: library_main +5749674 bytes (no line)
 10: library_main +5662462 bytes (no line)
 11: library_main +5664210 bytes (no line)
 12: library_main +5747910 bytes (no line)
 13: library_main +318776 bytes (no line)
 14: library_main +3253909 bytes (no line)
 15: library_main +3254363 bytes (no line)
 16: BaseThreadInitThunk +36 bytes (no line)
 17: RtlSubscribeWnfStateChangeNotification +1081 bytes (no line)
 18: RtlSubscribeWnfStateChangeNotification +1028 bytes (no line)

Running TvdL EDRW didn't cause a crash, but it did hang on the first rebuild.

[actiasluna]

Segmentation Fault 11… will post feedback. Crashed 3x… so this is still broken on my Mac

[actiasluna]

when using some of the shake-wiggle-mutate functions… cancel or stop doesn't work (intermittent), ESC will stop or cancel.

[kabubi]

can't see the player list too

[Timo van der Laan]

Shake takes forever, because it also shakes the big part.

[gitwut]

I ran Local Quake on it for a while but the progress was way too slow (probably because of shake). I switched to Enhanced DRW on small subsets of the protein but encountered a crashes almost immediately. I don't currently have the inclination, patience or time to coddle this type of puzzle along.