2669: Electron Density Reconstruction 138

Closed since 5 months ago

Summary

• Created: September 25, 2025
• Expires: October 09, 2025 at 18:00 UTC
• Max points: 100

Description

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. This puzzle has multiple copies in it of a part of the SARS-COV1 spike protein.

• Sequence: QKQIANQFNKAISQIQESLTTTSTALGKLQDVVNQNAQALNTLVKQINASVVNIQEEIDR LNEVAKNLNQKQIANQFNKAISQIQESLTTTSTALGKLQDVVNQNAQALNTLVKQISGIN ASVVNIQEEIDRLNEVAKNLNESLIDLQELQKQIANQFNKAISQIQESLTTTSTALGKLQ DVVNQNAQALNTLVKQINASVVNIQEEIDRLNEVAKNLNESLIDLQELNQKQIANQFNKA ISQIQESLTTTSTALGKLQDVVNQNAQALNTLVKQGINASVVNIQEEIDRLNEVAKNLNE SLIDLNQKQIANQFNKAISQIQESLTTTSTALGKLQDVVNQNAQALNTLVKQLINASVVN IQEEIDRLNEVAKNLNESLNQKQIANQFNKAISQIQESLTTTSTALGKLQDVVNQNAQAL NTLVKQNASVVNIQEEIDRLNEVAKNLNESLIDL

Top groups

• 1. Go Science 100 pts. 57,075
• 2. Contenders 63 pts. 57,070
• 3. Anthropic Dreams 37 pts. 57,023
• 4. L'Alliance Francophone 21 pts. 56,718
• 5. Australia 11 pts. 56,661
• 6. FamilyBarmettler 5 pts. 56,548
• 7. Void Crushers 2 pts. 56,007
• 8. Marvin's bunch 1 pt. 55,911
• 9. VeFold 1 pt. 55,773
• 10. Gargleblasters 1 pt. 55,517

Top evolvers

• 1. Bruno Kestemont 100 pts. 57,075
• 2. Anfinsen_slept_here 47 pts. 57,075
• 3. Bletchley Park 19 pts. 57,070
• 4. Galaxie 7 pts. 57,023
• 5. LociOiling 2 pts. 56,991
• 6. alcor29 1 pt. 56,960
• 7. BootsMcGraw 1 pt. 56,538
• 8. bravosk8erboy 1 pt. 56,259

Comments

[LociOiling]

This puzzle is a match for PDB 1WNC, "Crystal structure of the SARS-CoV Spike protein fusion core".

PDB 1WNC has six chains, each of which should contain the same sequence of amino acids. Each of the chains has a large section of missing residues in the middle, with more missing residues at the beginning of the chain. Three out of the six chains also have missing residues at the end.

Missing residues in the middle of a chain cause recipes like AA Edit 3.0 to see two chains instead of just one. So the six chains in 1WNC become 12 chains in Foldit.

Based on comparing the missing residues, chains A and B in the puzzle correspond to chain A in 1WNC, while puzzle chains C and D match 1WNC chain B.

The remaining chains of 1WNC follow the same pattern. Chain C is Foldit chains E and F , chain D is Foldit G and H, chain E is Foldit H and I, and chain F is Foldit J and K.

(Update: the previous version was a little confusing, tried to clarify how and why the 12 chains in Foldit match the six chains in the PDB.)

[spvincent]

What exactly are the problems with the published solution? The structure seems to match the electron density data pretty well.

[LociOiling]

Good question. I suspect the issue is that is has poor statistics. Kind of a moneyball thing.

Looking at the summary page for PDB 1WNC, the graph shows poor values for clashscore and Ramachandran outliers. Clashscore probably refers to clashing, and Rama outliers would probably be segments that fall outside the colored areas of the Rama map.

The rfree value is also on the low side. Rfree is a measure of how close the published solution is to the experimental data. I'm not sure how exactly that's measured, but it would probably be related to the density score in Foldit. (Guesswork there.)

[LociOiling]

The general Rama map (nothing selected) for the starting pose is kind of all over the place. Clicking on the outliers changes the color map a little, but doesn't really help for the most part. So that's a good picture of the Rama outliers.

Puzzle 2669 also starts with overall clashing at -11,487, which probably explains the clashscore results.

Google gives this AI summary of RSRZ:

RSRZ (Real Space R-value Z-score) outliers are residues in a protein structure that exhibit a poor fit to their corresponding electron density map, indicating potential errors in the atomic model.

So it's a little unclear how 1WNC has poor rfree, but a relatively good score on RSRZ outliers.

There's also a user guide to the wwPDB X-ray validation reports at wwpdb.org, that goes into more detail about what the validation statistics mean.

• Rama map for the starting pose of puzzle 2669.

[spvincent]

Tx Loci: I daresay that's what the issue is. I don't usually pay a great deal of attention to the Rama map I'm afraid: relying instead on the Backbone/Ideality scores.