Placeholder image of a protein
Icon representing a puzzle

450: Huge Unsolved Leishmania Parasite Protein

Closed since over 14 years ago

Advanced Overall Prediction

Summary

Created
August 26, 2011
Expires
Max points
100
Description

This one's going to be tough. Every attempt so far to solve this 244-residue protein has failed. There are no NMR models to use as starting points, so we are giving you 5 Zhang Server predictions to start from. Each time you reset the puzzle one of these 5 models will randomly be chosen. They are also in the Alignment Tool so you can use partial threading. Since this protein is so big we are giving you 20 days to work on it (for 200 global points) and we'll post as much help as we can in the puzzle comments.

Top groups

  1. Avatar for SETI.Germany 11. SETI.Germany 10 pts. 12,759
  2. Avatar for Natural Abilities 12. Natural Abilities 8 pts. 12,746
  3. Avatar for Penny-Arcade 13. Penny-Arcade 6 pts. 12,429
  4. Avatar for Mojo Risin' 14. Mojo Risin' 4 pts. 12,041
  5. Avatar for Nerdfighters 15. Nerdfighters 3 pts. 11,836
  6. Avatar for Czech National Team 16. Czech National Team 2 pts. 11,809
  7. Avatar for Eὕρηκα! Heureka! 17. Eὕρηκα! Heureka! 2 pts. 11,190
  8. Avatar for Team China 18. Team China 1 pt. 11,112
  9. Avatar for Crunching Family 19. Crunching Family 1 pt. 10,920
  10. Avatar for CMU-PITT Comp Bio 20. CMU-PITT Comp Bio 1 pt. 10,789
  1. Avatar for gramps
    1. gramps Lv 1
    100 pts. 13,973
  2. Avatar for steveB 2. steveB Lv 1 99 pts. 13,798
  3. Avatar for gdnskye 3. gdnskye Lv 1 97 pts. 13,713
  4. Avatar for mimi 4. mimi Lv 1 96 pts. 13,710
  5. Avatar for wudoo 5. wudoo Lv 1 94 pts. 13,651
  6. Avatar for auntdeen 6. auntdeen Lv 1 93 pts. 13,642
  7. Avatar for PRL 7. PRL Lv 1 91 pts. 13,631
  8. Avatar for Deleted player 8. Deleted player pts. 13,614
  9. Avatar for mat747 9. mat747 Lv 1 89 pts. 13,605
  10. Avatar for anthunk 10. anthunk Lv 1 87 pts. 13,602

Comments

beta_helix Staff Lv 1

but due to Foldit's success with the previously unsolved monkey virus protein (http://fold.it/portal/node/989012) we wanted to let you try out this big protein since every other method to solve it has failed.

Here are some tips to help your computer run Foldit faster:

After opening this puzzle for the first time (before even shaking or wiggling)
go to View and turn off "Show sidechains with clashes or exposeds"
Under View Sidechains: select "Don't Show (Fast)"
Turn off everything under View Options (especially "Show Clashes")

If you selected "Show Advanced GUI"
(available under the Menu->General Options button) then in the View Menu turn on "Fade GUI" (or even "Hide GUI" at times if you don't need it) and when you plan to run something for a while (even just shake or wiggle) select "Line" under View Protein. Then set it back to Cartoon when you want to manipulate the protein again.

Foldit can run faster if it doesn't have as much graphics to render, so when running recipes (or long wiggles/shakes) you can move the protein off of the screen (and get it back by pressing Q) and if you are on PC you can minimize the Foldit application to your taskbar and it won't render the graphics.

This puzzle will be tough, but once you get rid of the clashes (don't forget to use the Behavior Tab) you should be able to manipulate it.

Please post any other speedups that you know of!
Thanks and good luck!

Rav3n_pl Lv 1

I takes over 2.5 hours but:

Save 1 start points: -35170.683355334 - model 2
Stabilzed, fuzed and saved at: 11441.483779276
Save 2 start points: -35411.111519259 - model 3
Stabilzed, fuzed and saved at: 10998.760213389
Save 3 start points: -26984.75861465 - model 4
Stabilzed, fuzed and saved at: 11180.802313728
Save 4 start points: -67135.737279717 - model 5
Stabilzed, fuzed and saved at: 11487.285480836
Save 5 start points: -39946.359775718 - model 1
Stabilzed, fuzed and saved at: 11856.137023034

Looks like model 1 is scoring best at start.

beta_helix Staff Lv 1

This is the unsolved protein:
http://www.ncbi.nlm.nih.gov/protein/76363613?report=genbank&log$=protalign&blast_rank=1&RID=5K54MX9Z01S

Not much info about it, but luckily there is a homologous solved structure: 2pxx (http://www.rcsb.org/pdb/explore/explore.do?structureId=2pxx) it just wasn't close enough to solve the structure using Molecular Replacement.

You can see how that homolog was used by the Zhang Server for its 5 predictions (you'll also find secondary structure predictions, hydrophobicity predictions, different alignments to 2pxx and more) here: http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S79415/

Lastly, below are the secondary structure predictions (in sequence logo form) from the SAM-T08 server (since they don't agree perfectly with the Zhang server secondary structure predictions):

gramps Lv 1

This may be coincidence but I've recently found the setting to prevent Skype from starting up in the background when I log in, and recipe performance seems to have become friskier since (Win7). In general it's probably better to have as little other major S/W running as possible when folding, but I just have a feeling that Skype in particular is a real hog. Presumably things like streaming video and such would also affect folding performance.