2490: Refine Density Reconstruction 3

Closed since over 1 year ago

Summary

• Created: July 29, 2024
• Expires: August 08, 2024 at 18:00 UTC
• Max points: 100

Description

This is a protein we've given before in puzzle 2466, which was Reconstruction Puzzle 92, but now we have the Refine Density tool available to make folds even better! Learn about the new tool here: https://fold.it/forum/blog/new-tool-refine-density. This one should allow loading solutions from that puzzle.

• Sequence: MNCFEMLRCDEGLRLKIYKDTEGYYTIGIGHLLTKSPSLNAAKSELDKAIGRNTNGVITKDEAEKLFNQDVDAAVRGILRNAKLKPVYDSLDAVRRCALINMVFQMGETGVAGFTNSLRMLQQKRWDEAAVNLAKSRWYNQTPNRAKRVITTFRTGTWDAYKNC

Top groups

• 1. Anthropic Dreams 100 pts. 28,296
• 2. Go Science 68 pts. 28,099
• 3. Team China 44 pts. 28,004
• 4. L'Alliance Francophone 27 pts. 27,846
• 5. Contenders 16 pts. 27,828
• 6. VeFold 9 pts. 27,610
• 7. Marvin's bunch 5 pts. 27,521
• 8. Void Crushers 3 pts. 27,451
• 9. Australia 1 pt. 26,797
• 10. FamilyBarmettler 1 pt. 26,785

Top soloists

• 1. LociOiling 100 pts. 28,201
• 2. Punzi Baker 3 93 pts. 28,162
• 3. Galaxie 86 pts. 28,115
• 4. NinjaGreg 80 pts. 28,099
• 5. blazegeek 74 pts. 28,063
• 6. Zhang Ruichong 68 pts. 28,004
• 7. vs 63 pts. 27,930
• 8. Bruno Kestemont 58 pts. 27,921
• 9. grogar7 53 pts. 27,919
• 10. gmn 49 pts. 27,913

Comments

[NinjaGreg]

Just noticed that the scoring of the two puzzles is different by a factor of 2, so the prior results tend to overshadow the newer ones. Not sure if that was intended. Also, when you load a prior puzzle solution is does adjust the score downward, but then you are working with four strands, not the 2490 original two.

[LociOiling]

I was able to load solutions from puzzle 2466. The score I get in 2490 is a little less than half the 2466 score.

I'm not seeing a difference in the primary structure between 2466 and 2490. Here's what "print protein 2.9.4" reports for both:

segment count = 328
4 disulfide bridges found, segment pairs = 
3,97 9,164 167,261 173,328 
0 ligands
2 chains
chain A, type = protein, segments = 1-164, length = 164, mutables = 0
chain B, type = protein, segments = 165-328, length = 164, mutables = 0

Chain identification is still a work in progress in "print protein" and related recipes, so take that with a grain of salt.

I'll look at this again a little later, but right now the big question for me is why the scoring is so different in the two puzzles.

[LociOiling]

I should also mention that any 2466 solutions will be "local" solutions, meaning they live in C:\Foldit (or the equivalent), and not on the Foldit server.

Your 2466 solutions will appear on the right-hand side of the Open/Share Solutions dialog. You must have started Foldit for puzzle 2490 in the same C:\Foldit directory as when you played puzzle 2466.

In Open/Share Solutions, any "previous puzzle" solutions appear in blue, with "[prev. puzzle]" added to the solution title.

I make a final save when each puzzle finishes, and include the puzzle name in the description. That's helpful just in case a future puzzle allows loading a past solution.

[horowsah]

We'll have to look into the scoring difference and how best to handle it, but my initial guess is it comes from the way the maps are generated, which are quite different for these two different puzzles as for the original one it could be highly curated, and in the refine density puzzle it had to be made refineable.

[LociOiling]

The scoring difference doesn't seem to be a real problem, it's just a little confusing. It's only an issue if you expect the scores to be directly comparable between 2466 and 2490.

[Sandrix72]

For me in half of the cases after I push the Refine Density button, the whole program closes automatically. It has to be an annoying bug here!