Placeholder image of a protein
Icon representing a puzzle

2323: Electron Density Reconstruction 47

Closed since over 2 years ago

Novice Overall Prediction Electron Density

Summary

Created
June 15, 2023
Expires
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 structure has three copies in it, but they have rather different amounts of amino acids that are actually visible.

Sequence
MQSINFRTARGNLSEVLNNVEAGEEVEITRRGREPAVIVSKATFEAYKKAALDAEFASLFDTLDSTNKELVNR

Top groups

  1. Avatar for Go Science
    1. Go Science
    100 pts. 22,221
  2. Avatar for Anthropic Dreams 2. Anthropic Dreams 60 pts. 22,179
  3. Avatar for L'Alliance Francophone 3. L'Alliance Francophone 33 pts. 22,028
  4. Avatar for Contenders 4. Contenders 17 pts. 21,730
  5. Avatar for Australia 5. Australia 8 pts. 21,721
  6. Avatar for FamilyBarmettler 6. FamilyBarmettler 4 pts. 21,665
  7. Avatar for VeFold 7. VeFold 1 pt. 21,368
  8. Avatar for AlphaFold 8. AlphaFold 2 pts. 21,368
  9. Avatar for Marvin's bunch 9. Marvin's bunch 1 pt. 21,154
  10. Avatar for Gargleblasters 10. Gargleblasters 1 pt. 19,695
  1. Avatar for abiogenesis 31. abiogenesis Lv 1 4 pts. 20,749
  2. Avatar for rosie4loop 32. rosie4loop Lv 1 4 pts. 20,675
  3. Avatar for georg137 33. georg137 Lv 1 3 pts. 20,632
  4. Avatar for ProfVince 34. ProfVince Lv 1 3 pts. 20,429
  5. Avatar for wosser1 35. wosser1 Lv 1 3 pts. 20,103
  6. Avatar for roarshock 36. roarshock Lv 1 2 pts. 19,972
  7. Avatar for hada 37. hada Lv 1 2 pts. 19,879
  8. Avatar for Joanna_H 38. Joanna_H Lv 1 2 pts. 19,695
  9. Avatar for carxo 39. carxo Lv 1 1 pt. 19,591
  10. Avatar for rezaefar 40. rezaefar Lv 1 1 pt. 19,502

Comments

LociOiling Lv 1

This week's ED recon features something called Phd, for "Prevent Host Death". Phd works to suppress the toxic Doc. No idea what Doc stands for however. See PDB 3HRY.

There are supposed to be three more or less identical chains here. Recipes like AA Edit find more chains because some segments are spaced unreasonably far apart.

If you look at the 3D model for 3HRY, you'll see a dotted line between residues 65 and 66, which are far apart. Take a look at segments 65 and 66 in Foldit. The dotted line is replaced by two really long segments, and both segments have ideality of about -310,000 at the start. Similar problems occur in the second and third chains.

LociOiling Lv 1

Looking at what should be chain C, segments 129 and 130 also start out very long, although their ideality is only about -38,000 each.

In 3HRY, there are five missing residues reported at that point. If you look at 3HRY in PyMol or JMol, , you'll see a gap between segments 8 and 14 of chain C, allowing space for the missing residues. Foldit has no gap, with 129 (8C) bonded to 130 (14C).

We've seen similar problems with missing residues in other puzzles. It seems like the Foldit puzzle isn't following the experimental results. So we're stuck trying to cover segments with density that belongs to a different sidechain.

rosie4loop Lv 1

I've posted a bug report of a more detailed investigation of the problem in https://fold.it/forum/bugs/puzzle-2323-issues-with-missing-densities-in-alternate-locations-chain-a-and-false-residue-connectivity-chain-c

The problem in chain A is due to the different amount of segments visible in alternate conformations of residues 58-67, that 66-67 are only visible in alternate location B, but in the game only one alternate location of each residues have been extracted.

Actually densities connecting segments 58-66 in altloc B are missing as they cut the densities around the protein.

rosie4loop Lv 1

Alternate locations are atoms visible in multiple positions in xray density, usually assigned with occupancies lower than 1. In 3hry chain A residues 58-67, altloc A and B have been assigned 50% occupancies, even though 66-67 are only visible in altloc B.

LociOiling Lv 1

Thanks rosie4loop, I'll look at the bug report. Meantime, here's the side-by-side, showing 3HRY in Jmol on the left, and Foldit on the right. It seems like Foldit should have the same gap seen in the PDB model. Or else it could somehow cram in the missing segments and let us sort them out….

horowsah Staff Lv 1

Yeah, this is something that can be hard to spot when posting puzzles- the issue can be seen in the pdb# where there's a gap. There's a way when posting we can help alleviate these issues a bit if we spot them ahead of time, but sometimes they slip through. Thanks for pointing out- might be a good one to repost at a later date with that fix and see how players change what they do.

LociOiling Lv 1

This puzzle was puckered due to missing residues.

The recipe Pucker Picker 3.0 RC 1 detected these puckers:

2323: Electron Density Reconstruction 47
3  puckers found!
pucker 1 (ideality), segments 65-66 (protein), distance = 27.435, ideality = -310889.843, -310888.017
pucker 2 (distance/ideality), segments 118-119 (protein), distance = 4.661, ideality = -5832.645, -5832.211
pucker 3 (ideality), segments 129-130 (protein), distance = 9.193, ideality = -38437.355, -38437.401