Icon representing a puzzle

2191: Electron Density Reconstruction 12

Closed since 24 days ago

Novice Overall Prediction Electron Density


August 25, 2022
Max points

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.



Top groups

  1. Avatar for Gargleblasters 100 pts. 18,901
  2. Avatar for Anthropic Dreams 2. Anthropic Dreams 68 pts. 18,891
  3. Avatar for Go Science 3. Go Science 44 pts. 18,874
  4. Avatar for Void Crushers 4. Void Crushers 27 pts. 18,828
  5. Avatar for L'Alliance Francophone 5. L'Alliance Francophone 16 pts. 18,787
  6. Avatar for Contenders 6. Contenders 9 pts. 18,758
  7. Avatar for Marvin's bunch 7. Marvin's bunch 5 pts. 18,725
  8. Avatar for FamilyBarmettler 8. FamilyBarmettler 3 pts. 18,492
  9. Avatar for WISE 380 Spring 21 A 9. WISE 380 Spring 21 A 1 pt. 18,341
  10. Avatar for Australia 10. Australia 1 pt. 18,152

  1. Avatar for Enzyme
    1. Enzyme Lv 1
    100 pts. 18,901
  2. Avatar for grogar7 2. grogar7 Lv 1 95 pts. 18,876
  3. Avatar for LociOiling 3. LociOiling Lv 1 91 pts. 18,870
  4. Avatar for Sandrix72 4. Sandrix72 Lv 1 86 pts. 18,868
  5. Avatar for jeff101 5. jeff101 Lv 1 81 pts. 18,846
  6. Avatar for Galaxie 6. Galaxie Lv 1 77 pts. 18,843
  7. Avatar for Timo van der Laan 7. Timo van der Laan Lv 1 73 pts. 18,828
  8. Avatar for Punzi Baker 3 8. Punzi Baker 3 Lv 1 69 pts. 18,823
  9. Avatar for ucad 9. ucad Lv 1 65 pts. 18,822
  10. Avatar for Aubade01 10. Aubade01 Lv 1 62 pts. 18,822


jeff101 Lv 1

<pre>When Tabbing on different segments in this puzzle, I noticed that asp65 says PDB# A181 while ala66 says PDB# A183. Does this mean that residue PDB# A182 is missing, perhaps due to a blurry section in the Electron Density Cloud? If so, what amino acid is PDB# A182? Should we treat asp65 and ala66 as if they are far enough apart for PDB# A182 to fit between them? Should we instead treat asp65 and ala66 as if they are directly covalently-bonded to each other via a peptide bond?

If we need to leave space for the missing
residue PDB# A182, perhaps this puzzle should
break up the protein into two chains (one for
segments 1-65 and another for segments 66-96).
Maybe the 2 chains could be connected to each
other with a cut band or some other constraint.
What do you think?


jeff101 Lv 1

This puzzle's protein contains 3 cysteines.
Should our solutions have any disulfide bonds?

bkoep Staff Lv 1

Yes it looks like we are missing a residue in the starting structure. Good find!

I'm not exactly sure how this happened; I'll have to take a look. In the published structure this residue is modeled with "zero occupancy," which is kind of like modeling that residue as a placeholder even though there is no density to suggest how the residue is folded.

jeff101 Lv 1

If there is a region in the Electron Density Cloud 
with "no density to suggest how the residue is folded",
instead of giving us a protein that is missing a residue 
there, why not give us the entire protein and instead 
change the Electron Density Cloud some? 

Would replacing the bad region with a constant density 
help? Some ideas for the constant density could be zero, 
the average density in the cloud, or the maximum density 
in the cloud. Zero would work best for generally empty 
regions, and large positive values would work best for 
tightly-packed interior regions of the protein. 

Another idea would be to approximate the overall 
Electron Density D as a spherical Gaussian in 3D, like: 
D=density=A*exp(-k*((x-xo)^2 + (y-yo)^2 + (z-zo)^2))
where A and k are positive constants, and xo,yo,zo is 
the approximate center of the Electron Density Cloud. 
With a good enough approximation, one could then 
replace the bad region with the density D given by 
the formula above.