1221: Disordered Electron Density: Round 1

Closed since almost 10 years ago

Summary

• Created: April 17, 2016
• Expires: April 24, 2016 at 23:00 UTC
• Max points: 100

Description

NOTE: Because many players have had trouble playing this puzzle, the puzzle has been made worth zero points, and it will not contribute to player rankings. However, this puzzle is still of unique scientific interest, and we encourage players to work on it if they are able!



This is a big one! This electron density puzzle features two proteins: the first is a large well-ordered helical protein that is mostly frozen in place. This binder protein has two flexible loops that bind a second, mostly disordered, substrate protein. The substrate protein is provided here as an extended chain, and is completely flexible. We know that the substrate protein is bound to the binder protein somewhere in this crystal structure. But it seems that the substrate is able to adopt several different shapes, making the electron density very difficult to interpret. We’re asking Foldit players to help us find the ways that the substrate protein can bind to the binder protein! Fold up the extended chain and dock it next to the flexible loops of the binder to improve your score!

Top groups

• 1. Go Science 100 pts. 10,353
• 2. Contenders 77 pts. 10,328
• 3. Beta Folders 58 pts. 10,313
• 4. Anthropic Dreams 43 pts. 9,849
• 5. BOINC@Poland 31 pts. 9,820
• 6. Gargleblasters 22 pts. 9,755
• 7. L'Alliance Francophone 15 pts. 9,282
• 8. HMT heritage 11 pts. 8,863
• 9. Italiani Al Lavoro 7 pts. 8,661
• 10. Deleted group pts. 8,308

Top soloists

• 1. dcrwheeler 100 pts. 10,450
• 2. Bruno Kestemont 97 pts. 10,353
• 3. Bletchley Park 94 pts. 10,328
• 4. bertro 91 pts. 10,313
• 5. pauldunn 88 pts. 10,257
• 6. gloverd 85 pts. 10,253
• 7. mirp 82 pts. 9,936
• 8. mimi 79 pts. 9,919
• 9. jamiexq 76 pts. 9,849
• 10. kitek314_pl 74 pts. 9,820

Comments

[bendbob]

Just wondering (the old software QA guy in me is asking):

1. can you move the quicksaves to a temporary fold
2. start game
3. move the quicksaves back to the correct location
4. attempt to open a quicksave

Will the client crash now?

[inkycatz]

It would be interesting to see the results of this if anyone wants to try it out and report back.

[Bruno Kestemont]

I loose my best solution rank 5 after a computer stop.

Loci's solution doesn't work for me.

log says:
CRASH: 447652

When trying to open any new Foldit client, windows7 says
"The game has previously not started up corectly.If this problem persists, please install the latest version"

Completely deleted map foldit\puzzles\002002315

Opening: ok, a rest puzzle :(

Trying to load a saved solution: crash

[TomTaylor]

Followed your recommendations. I even shut down my PC and the first thing I did once up again was start foldit. Still crashed. I seriously doubt it's a memory issue.

I have a one year old i7 3.6 GHz, 12 GB ram PC running Windows 10. Unless foldit needs more than 10 GB of memory it should be fine. Other players have commented about having less powerful PCs and being able to work on the puzzle.

[LociOiling]

@bendbob, in this case, the puzzle was crashing upon loading. I hadn't run any scripts, so I didn't have any quicksave files in the puzzles directory.

All I had was four autosaves:

autosave.ir_solution
autosave-best.ir_solution
autosave-creditbest.ir_solution
autosave-recentbest.ir_solution

I saved the files from another crashy client and took a look.

All were about 120 KB, slightly smaller than the ones from a client running on the same puzzle. The files looked reasonable on inspection. They are most binary, with a few eyecatchers such as "VRFY" near the end. I'm assuming "VRFY" is similar to the "verify" keyword found in various other foldit files.

I tried deleting the files one-by-one in the order shown. The client wouldn't start until all four were gone.

I did a shake and wiggle once the client was running and compared the new autosave files to the old ones. I don't see any obvious differences in the overall structure. It appears autosave.ir_solution gets updated from time to time, even with the client sitting idle. There is some move information (references to ActionGlobalMinimize, ActionRepack) near the end of the new file, before the PHAS, ALGN, SCOR, and VRFY eyecatchers. Bottom line, the bad files aren't obviously corrupted or truncated.

I'll be glad to share the "bad" files with the developers.

[frood66]

this puzzle is currently not fit for purpose.

Bonus science chat is after it's closure.

I suggest the puzzle is removed and fixed - then we can have the science chat. Then the puzzle can be reposted when it's fixed. This has to be more productive than what we have now?

[LociOiling]

First, there was an omission in my previous post, the saved solutions are under:

foldit\puzzles\002002315*userid*trackname*.ir_solution

The previous post left out trackname, which is "default" or the name of the track you're using.

The next part is a long sad story.

I tried stopping a client to share. Of course, the client hung due to yesterday's network outage. Anticipating this problem, I was running the procmon utility, but unfortunately it didn't give me a good idea of where the hang is. Also, I saw some Norton Security activity on the process, so I need to see what that's about. (I thought I had excluded the foldit directories from Norton's scanning, but I guess not.) I do have the procmon log if anyone's interested.

After closing the hung client, it of course crashed on restart. Deleting the autosave files in the above directory did not help. I had to delete the quicksave files as well to restart the client.

After restarting, I tried copying the quicksave files back in. There were quite a few, as the client had been running TvdL Enchanted DRW 2.8.1. Foldit apparently only checks these files at startup, so things like "save.Quickload(50)" didn't change the pose of the protein.

After closing the client again, it of course crashed again on startup. So it appears that the work done on this client has been lost, unless I can figure something out.

[jamiexq]

Why leave a puzzle up that is so obviously broken - even if it is now for zero points?

[alcor29]

74 people have been able to log in and do some work on it. Out of that maybe 15 or 25 have been able to take it pretty far. The rest are probably locked out after crashing trying to make a new track, share etc. In fact since the scores don't seem to have moved, it is possible that everyone is now locked out. So why keep it active?

[Bruno Kestemont]

I'm ok with the team decision to let this puzzle running for 0 points.

My investment would not be completely lost and no player would be penalized.

Moreover, I find it quite challenging to try a puzzle in one unique run at a time (no share, no parallel track, no stop). A kind of pure "hand fold" or "devprev-type" round without points.

A good way to train for this new type of puzzle.