I modelled Top7 in POV, and I have found that it has in it,

ATOM    248  N   LEU A  36      11.213   6.021  16.730  1.00 59.14           N
ATOM    249  CA  LEU A  36      10.469   4.828  16.342  1.00 59.14           C  
ATOM    250  C   LEU A  36      11.321   3.923  15.466  1.00 59.14           C  
ATOM    251  O   LEU A  36      11.229   2.702  15.544  1.00 59.14           O  
ATOM    252  CB  LEU A  36       9.217   5.227  15.613  1.00 61.28           C  
 
and all other LEU have 8 atoms (top7) 

ATOM    201  N   LEU A  29       8.883  15.919  15.895  1.00 97.20           N
ATOM    202  CA  LEU A  29       8.286  14.820  15.147  1.00 97.20           C
ATOM    203  C   LEU A  29       9.475  14.012  14.649  1.00 97.20           C
ATOM    204  O   LEU A  29       9.456  12.784  14.643  1.00 97.20           O
ATOM    205  CB  LEU A  29       7.493  15.350  13.947  1.00 66.59           C
ATOM    206  CG  LEU A  29       6.452  14.423  13.296  1.00 66.59           C
ATOM    207  CD1 LEU A  29       6.087  14.961  11.915  1.00 66.59           C
ATOM    208  CD2 LEU A  29       6.997  13.012  13.162  1.00 66.59           C

It looks like ASN (top7)

ATOM    234  N   ASN A  34      12.719   9.404  14.389  1.00 69.48           N
ATOM    235  CA  ASN A  34      14.011   8.779  14.631  1.00 69.48           C
ATOM    236  C   ASN A  34      13.870   7.727  15.722  1.00 69.48           C
ATOM    237  O   ASN A  34      14.350   6.605  15.584  1.00 69.48           O
ATOM    238  CB  ASN A  34      15.039   9.844  15.038  1.00 46.71           C

Am I missing something (I did rotate an ASN with the 5 atom LEU and they looked the same to me)?

My 3D maths is non-existent, so I can't tell if atoms have a different relationship, if this is correct what happened to the other atoms?

Edit: a couple of times to try to get the format readable, and because I hit 4 instead of 5

Hi Hugo,

Xray crystallography relies on the fact that each asymmetric unit in a protein crystal is identical to the next. In reality, there are often some atoms within the protein that are disordered and are in different conformations in each assymetric unit. These atoms cannot be resolved within the electron density, and are therefore not included in the final model. It is better to truncate amino acids to the part that can be modeled with certainty than to include atoms that are unresolved in the density and give the impression that they are well ordered when they are not.

All amino acids (except Gly) are identical up to the C-beta (CB) atom. The differences become apparent in the atoms that are further from the backbone.

Cheers,
V.

---

Thanks Vanita.

I was beginning to think I was asking a really stupid question.

Edit: If I understand this correctly X-ray crystallographers, can tell that (in top7's instance) it's definitely not ASN, and can somehow tell that it is LEU, they just don't know the correct positions of the atoms.

Please disregard the previous edit.

I'm sure that they have would have access to the genetic code and therefore the protein sequences, too :)

you got it right ;-)
And it was a really insightful question to begin with. I'm always impressed at how all of you guys are really looking at the details.
V.

I'm sure that they have would have access to the genetic code and therefore the protein sequences, too :)

---

I know it's been over a year since this thread began, but I thought this would be a good place to ask another TOP7-related question. Is there a coordinate file for the lowest energy decoy of TOP7 available? Even though graphics showing the superposition of TOP7's best decoy and its crystal structure are available in a lot of places, I'd like to make a visualization of it myself in PyMol if possible.

---

I believe this is the one you want here:
1QYS
You can display this with any number of viewers available on that website, or you can import the atomic configuration of the model into the tooling of choice.

Edit:
Or, you could download the game, and open an older puzzle labeled as puzzle 61, and view it, and manipulate it in 3D with the game's controls. ...but, unless you are good with the game :), this is not the native structure of TOP7.

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/

I believe this is the one you want here:
1QYS
You can display this with any number of viewers available on that website, or you can import the atomic configuration of the model into the tooling of choice.

Edit:
Or, you could download the game, and open an older puzzle labeled as puzzle 61, and view it, and manipulate it in 3D with the game's controls. ...but, unless you are good with the game :), this is not the native structure of TOP7.

1QYS is the x-ray crystal structure, onto which I'd like to superposition the conformation for TOP7 predicted by Rosetta (like I did for Rosetta's first atomic-level accuracy prediction in CASP here). The reason I'm looking for the predicted atomic coordinates is because it's one of Rosetta's significant accomplishments, and the superposition would be a good feature on the Rosetta@home Wikipedia article I'm working to improve.

The Wikipedia administrators are strict about licensing/copyright issues (removing any images without licensing information), so I don't think I would be able to simply crop and upload a picture like the one here, for example, without its creator licensing it under GFDL and/or Creative Commons and letting wiki admins know. If I could create a similar superposition with the predicted and determined coordinates myself, then I would license it as freely available for non-commercial use (which it effectively already is without the license) and add it to the R@H Wikipedia article.

Regarding puzzle 61 in FoldIt, I don't imagine it's possible to get the TOP7 structure predicted by Rosetta. If it is, that'd be awesome -- but I'd like to have the graphic accurately reflect the ~1.2 Å RMSD between TOP7's Rosetta-predicted and x-ray crystal structures. Thanks for the pointer though :-)

---

Brian Kuhlman kindly emailed me the model. I posted it at:

https://boinc.bakerlab.org/rosetta/rah_misc/top7_model.pdb

Thanks! I posted the picture below here.

---