beta_helix's picture
User offline. Last seen 18 hours 25 min ago. Offline
Joined: 05/09/2008
Groups: None
PSIPRED results
Joined: 09/24/2012
Groups: Go Science
Translated:

LLLLLLLLLLLLLLLEELLLLLEEEELLHHHHHHHHHHHHLLLLLLLLLLL
LLLLLLLLLLLLLLLHHHHHHHHHHHHHHHHHLLLLLEEEEEEEEELLLLLLEEEEEEEEEEL

Susume's picture
User offline. Last seen 21 hours 41 min ago. Offline
Joined: 10/02/2011
floppy tyrosine?

I am used to the idea that long blue sidechains flop around in solution and may not show up well in the cloud, but in 1606 there seems to be an orange tyrosine that has two preferred positions, both of which appear clearly in the cloud. 112 TYR has a primary pocket for the sidechain that points toward the sidechain of 91 LYS. If the ED threshold is high, this is the only pocket visible. But if you lower the threshold a bit, a secondary pocket becomes visible, about 90 degrees off from the first, pointing toward the sidechain of 89 LEU. This pocket, like the other one, even has a tip for the tip of the TYR to fit into. I have one track where 112 TYR has settled into the primary pocket (density score 35.4, segment score 49.3), and another track where it has settled into the secondary pocket (density score 28.4, segment score 58.0). One of my teammates also reports a high-scoring TYR 112 that has settled in the less visible pocket.

Is it realistic that a tyrosine would have two clear preferred positions, or do you think this is just an artifact in the cloud?

Susume's picture
User offline. Last seen 21 hours 41 min ago. Offline
Joined: 10/02/2011
Solve ED with ensemble of positions?

In NMR (as opposed to crystallography) it seems common to take the average among an ensemble of conformations and call that the solution. In this case, since the two positions for the tyrosine are visibly discrete, one might argue that two overall conformations, one with 112 tyr in position A and the other in position B, would constitute an objectively better solution to the ED map than the average between them.

This begs the question of whether one could propose an ensemble of conformations as the best solution to an ED map, rather than a single conformation. Would the measures typically used to assess the goodness of a solution still make sense? Would protein scientists accept such an ensemble as a valid solution?

beta_helix's picture
User offline. Last seen 18 hours 25 min ago. Offline
Joined: 05/09/2008
Groups: None
Consulting with our expert experimentalist:

That is indeed interesting!
It’s a bit hard to visualize it without seeing pictures, but this sounds like a fairly typical case of alternate conformations.
While this is pretty common in crystal structures of reasonably high resolution, typically cryo-EM maps aren’t high enough resolution to see them... but it wouldn't be surprising if some of the more recently solved ones are high enough resolution to find them (on rare occasions).

Susume's picture
User offline. Last seen 21 hours 41 min ago. Offline
Joined: 10/02/2011
pics

1606 TYR 112 Pocket #11606 TYR 112 Pocket #1
1606 TYR 112 Pocket #21606 TYR 112 Pocket #2

Joined: 09/24/2012
Groups: Go Science
Score table

...............Pocket #1 .........../ Pocket #2

Total Score:......15840............./........16026
Tyr 112 score:.......51.1 ........../...........59.1
Ideality:.............-5.5........../...........-3.5
Backbone:..............5.9........../...........3.1
Sidechain:...........-19.1........../.........-14.1
Density:...........36.0........../..........29.3

So Foldit would prefer Pocket 2, which explains why I'm not satisfied with my best scoring solution (recipes tend to "destroy" the hand folded good density match).

I wonder if it's good to keep a parallel "best density score" solution and share it to scientists.

A question to Susume: using the new rotamer tool, which one is of higher probability?

beta_helix's picture
User offline. Last seen 18 hours 25 min ago. Offline
Joined: 05/09/2008
Groups: None
Consulting again with our expert experimentalist:

At this resolution it’s not easy to tell for sure, but they do look like they very could well be alternate conformers!

Nice catch, Susume!

This leads to the question: which state does Rosetta seem to prefer?

Get Started: Download
  Windows    OSX    Linux  
Windows
(7/8/10)
OSX
(10.7 or later)
Linux
(64-bit)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Search
Only search fold.it
Recommend Foldit
User login
Soloists
Evolvers
Groups
Topics
Top New Users
Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons