Placeholder image of a protein
Icon representing a puzzle

1848: Revisiting Puzzle 126: Ethanolamine Utilization

Closed since over 2 years ago

Novice Overall Prediction


June 05, 2020
Max points

This is a throwback puzzle to the early days of Foldit. This protein allows bacteria to metabolize ethanolamine and use it in constructing cell walls and cell membranes. The protein is modeled here in reduced form, so no disulfides are expected to form. We are revisiting old Foldit puzzles so we can see how useful the recent additions to the game have been and to provide newer players with puzzles that are still scientifically relevant.



Top groups

  1. Avatar for Beta Folders 100 pts. 10,921
  2. Avatar for Anthropic Dreams 2. Anthropic Dreams 76 pts. 10,784
  3. Avatar for Contenders 3. Contenders 56 pts. 10,722
  4. Avatar for Go Science 4. Go Science 41 pts. 10,688
  5. Avatar for Void Crushers 5. Void Crushers 29 pts. 10,534
  6. Avatar for L'Alliance Francophone 6. L'Alliance Francophone 20 pts. 10,488
  7. Avatar for Gargleblasters 7. Gargleblasters 14 pts. 10,469
  8. Avatar for Hold My Beer 8. Hold My Beer 9 pts. 10,330
  9. Avatar for Marvin's bunch 9. Marvin's bunch 6 pts. 10,266
  10. Avatar for Chem Eng Thermo 10. Chem Eng Thermo 4 pts. 9,661

  1. Avatar for LociOiling
    1. LociOiling Lv 1
    100 pts. 10,921
  2. Avatar for grogar7 2. grogar7 Lv 1 98 pts. 10,784
  3. Avatar for BootsMcGraw 4. BootsMcGraw Lv 1 94 pts. 10,720
  4. Avatar for dcrwheeler 5. dcrwheeler Lv 1 91 pts. 10,692
  5. Avatar for KarenCH 6. KarenCH Lv 1 89 pts. 10,684
  6. Avatar for Xartos 7. Xartos Lv 1 87 pts. 10,658
  7. Avatar for mirp 8. mirp Lv 1 85 pts. 10,657
  8. Avatar for Bruno Kestemont 9. Bruno Kestemont Lv 1 83 pts. 10,611
  9. Avatar for malphis 10. malphis Lv 1 81 pts. 10,545


marsfan Lv 1

These puzzles start out with a preliminary fold performed. Is that just a general guideline, or is it known that this protein folds to somewhere around that shape?

LociOiling Lv 1

Sometimes the revisiting puzzles are close to the native state, but sometimes they are farther away.

All the revisiting puzzles have at least one published solution, and some have a long list of solutions. You can always compare the starting pose to the published solution, as described below.

For the close ones, small differences can matter. For example, I think on the last ice binding revisit, puzzle 1845, the key to success may have been the "omega" backbone angle between segments 38 and 39. You can get close to the native if they're in the normal "trans" configuration, but I suspect they should be "cis" instead, with both sidechains more or less on the same side.

For other revisits, the starting pose may be farther away, and you can have major differences like sheets being out of order. There's a a whole series of "Rosetta decoy" revisits which mostly fall into this category. (For some reason, the Rosetta decoy puzzles were skipped in the last revisiting cycle.)

There are also some revisiting puzzles where the best Foldit results seem to be far from the published native protein. Not clear what's going on with those, but some of the smaller puzzles are like that.

Here's the scoop on this week's puzzle:

The protein in this puzzle matches two PDB entries. You can look at the solutions at, which includes an online protein viewer that's somewhat similar to Foldit. The matching entries are 2HD3 and 2Z9H, and both have multiple copies of the protein, which complicates viewing a little. For Foldit, we're working with a single copy of the protein, which in theory should fold into its native shape, and then bond to another copy of itself.

spmm Lv 1 - this is a free version of Pymol for educational purposes so you have to register. Other sites may also have versions.
The last one I had stopped working on the Mac when Apple moved to 64 bit - not sure if they have updated the edu version yet.

Why bother? Pymol allows you to display side chains which as you can see from Loci’s comment above, can be very important in strand orientation and points for the revisiting puzzles. You can also view a single chain in an assembly. Pymol is a professional tool and can take a bit of getting used to.

IMol viewer lite works well on an iPad - just type the PDB number in and it will display - limited choice of display options but quite useful. Side chains can be viewed as well.
There may be more recent tools of course for all platforms.

Mikisp Lv 1

VMD is also a very good professional tool to construct and view all proteins and show hydrogen bonds, of course. But it isn't very easy to use. Perse is an other viewer for Ipad.

Susume Lv 1

Is there a way to purposely make a cis bond in foldit? One could cut both sides of an amino acid and roll it 180 degrees around its backbone axis - but that risks making both ends of it cis, and wiggle is likely to fight with it as you try to stabilize. I suppose one could try very strong bands to space to try and force the sidechain to the other side of the backbone. Is there a better way?

jeff101 Lv 1
gives bond angles and lengths for cis & trans
peptide bonds. You could use these diagrams
and some trigonometry, like the law of sines
or the law of cosines
to find distances between certain atoms 
(like between 2 adjacent alpha-carbons), 
and then use bands to enforce these distances.
Some of my notes give Ca-Ca distances of
3.79 A for trans & 2.91 A for cis. 

Perhaps a more thorough google image search
would find a figure with more distances on it.