The structure of this protein has already been solved and published, but close inspection suggests that there are some problems with the published solution. We'd like to see if Foldit players can use the same electron density data to reconstruct a better model. There are several chains in this protein, some are the same, some are different from each other. One of them is just a short peptide. Given its large size, the Trim tool may be very much necessary to make progress!
It'd be nice if the gaps were treated as gaps when setting up the puzzle… if the sequence must be continuous within the same chain in foldit, use a different chain ID.
As a player, here are the recipes I found useful as a workaround, for the purpose of scoring.
Recipe SeletoPro v1.2 automatically split the chains for selection which is quite useful here to inspect the structure.
To get rid of the jumps automatically, quickfix is always my favourite recipe. Good for scoring, still the false connection is artificial.
It's currently too broken for me to run AA edit on. The sequence on the puzzle description blastps to some combination of arrestin and clathrin heavy 1.
Hmmm, rebuild doesn't seem to give any consideration to fixing the bond length whatsoever. Gotta manually select, do medium-strength backbone wiggle to normalize lengths, then rebuild.
lmao, just saw one of the bad idealities being not bond-length related: it was a flipped His. Shake one iteration before you go, people.
It's currently too broken for me to run AA edit on. The sequence on the puzzle description blastps to some combination of arrestin and clathrin heavy 1.
Hmmm, rebuild doesn't seem to give any consideration to fixing the bond length whatsoever. Gotta manually select, do medium-strength backbone wiggle to normalize lengths, then rebuild.
lmao, just saw one of the bad idealities being not bond-length related: it was a flipped His. Shake one iteration before you go, people.
I suggest extending the deadline for this puzzle.
With or without the trim tool, I've the impression that the computer speed is here determinant.
The trim tool might speed up time when hand folding.
But using recipes for repetitive tasks, i'm not sure there is a difference between working successively on trimmed parts or on the all protein.
We are sorry that it took so long for us to respond to this bug (unfortunately this occurred over the Thanksgiving holidays).
We are looking into preventing this from happening again and thank you for your patience!
@Sandrix72 We can indeed look into that, thanks for the suggestion.
After some detective work, it appears this puzzle is a match for PDB 3GD1.
Visually, the puzzle has three large protein chains, and a fourth 8-residue peptide.
In 3GD1, chains A and B started out with the same sequence. In 3GD1, these chains are identified as C and E, "Beta-arrestin-1".
Chain C in the Foldit puzzle is identified as chain I in 3GD1, "Clathrin heavy chain 1".
The puzzle's small chain D is chain Z in 3GD1, just "clathrin".
The segment information window reveals the chain and residue number from the PDB. For example, segment 1011 shows it was "Z 378" in the PDB. (Unfortunately, there's no Lua access to this information.)
The authors of 3GD1 were just being creative in naming their chains. There's no PDB requirement to start with chain A and then move on to chain B. So 3GD1 has chains C, E, I, and Z, but no others.
Aside from the size, the main problem in Foldit is that 3GD1 has many sections with missing residues, identified with "REMARK 465" records in the PDB file, stating "the following residues were not located in the experiment". The first two chains have the most holes. As Artoria2e5 mentions above, the puzzle simply connects the residues on either side of a range of missing segments. This approach results in long straight two-segment "straws" with terrible ideality scores.
For PDB chain C (our chain A), the missing residues are 1-2, 66-72, 91-93, 179-180, 332-338, and 357-383.
For PDB chain E (our chain B), the missing residues are 1-7, 13-18, 31-32, 43-54, 67-70, 83-97, 134-137, 150-159, 179-180, and 282-283.
The missing residues at the beginning don't cause a problem, but the ones in the middle do. The first two chains both end with the sequence "FEDFAR", which is residues 386-393 in the PDB. Gaps like 357-383, meaning 27 missing residues, tend to cause the biggest problem, since Foldit just connects 356 to 384. The missing 27 residues were there for a reason, and their absence means a lot of other things will end where they don't belong.
The good news is that chains C and D (known as I and Z in the PDB) don't have any missing residues, so they're a 100% match to what we have in Foldit.
@Joanna_H the fix for crashing may be to rename or delete some files.
Unfortunately, this can get a little technical. I'm assuming Windows here, but the details are similar for Mac and Linux. The suggestions here are easiest with Windows Explorer, but a command prompt is another option if you don't mind some typing.
In your Foldit folder (c:\Foldit or what have you), there's a puzzles folder (for example, c:\Foldit\puzzles).
The puzzles folder in turn has a folder for each puzzle you've played.
For puzzle 2383, the folder is named "2013749". That's the "node number" of the puzzle, which appears in the address of this page: https://fold.it/puzzles/2013749. The node number is padded on the left with zeroes.
So you should have C:\Foldit\puzzles\0002013749. Under that folder, you'll have one for your userid, so C:\Foldit\puzzles\0002013749\userid, where userid is your user "node number", again padded with zeroes.
Then under C:\Foldit\puzzles\0002013749\userid, there'll be a folder called "default", plus additional folders if you use the tracks feature. Each track you create gets a folder here. I'll assume you were in the "default" track, since Foldit crashes immediately. Otherwise, you'd start Foldit, but then it would crash when you changed tracks.
One option for fixing the crash is to selectively rename or delete files from "default". I think the most recent state is saved in a file called "autosave.ir_solution". This is most likely where the problem is. Try renaming that file to "autosave.ir_solution.old", then restart Foldit.
Another variation is to rename "default" folder to "default_old". Try that if renaming "autosave.ir_solution" doesn't work. After renaming, try starting Foldit again.
Assuming one of these methods works, you won't have your most recent work any more, but that solution was causing a crash. You'll still have any manual saves you did (using ctrl+s or ctrl+o).
If you used quicksaves, they'll be available in files like "quicksave99.ir_solution" for quicksave slot 99. Under Open/Share Solutions (ctrl+o), if you check "show auto and quicksaves", you'll see things like "Quicksaved Soloist Solution 99" in the list. (If you renamed the "default" folder above, you'll need to copy the files from "default_old" to the new "default", and you may need to restart to get the updated list.)
