Make viewer for saved/shared structures.

Case number:845833-992668
Topic:Game: Display
Opened by:jeff101
Status:Open
Type:Suggestion
Opened on:Tuesday, May 15, 2012 - 04:24
Last modified:Friday, October 25, 2013 - 21:59

It seems like Foldit saves files on my computer with names like
puzzle_992432_time_1335080575.ir_solution containing protein structures.
Is there a standard viewer like Rasmol that can read these files and display
the protein structures stored in them?
If not, perhaps you should include with Foldit
a viewer utility program outside of the present Foldit GUI
that will read these *.ir_solution structure files
that Foldit saves on a player's hard drive.

By the way, why doesn't Foldit output saved structures as *.pdb or *.ent files?
Doing so would let users do a wide variety of things outside of the Foldit GUI.
It would also make it easier to do comparisons among structures.
Perhaps include with Foldit a program that will convert
back and forth between *.ir_solution and *.pdb or *.ent files.

It would also be nice to have within the Foldit GUI
little snapshots of each saved/shared structure in some standard format,
perhaps all colored and rendered in a way that the local user specifies.
This will help folks more easily compare different structures
without having to load and view each one separately in the Foldit GUI, one at a time.
If a user could rotate these little images too in 3D space, that would be even better.
Barring user-controlled 3D rotation, you could at least make little snapshot movies of
the protein viewed from different angles, and have these playing when the menu of
saved/shared structures comes up. Many Nintendo games (for example, Mario Kart)
have rotating 3D objects in menus where players select their characters or vehicles,
for example. Why not have Foldit do the same?

(Tue, 05/15/2012 - 04:24  |  6 comments)


Joined: 04/19/2009
Status: Open » Closed

Duplicate to another long standing debate... Here is just one of the feedbacks that discussed this issue: http://fold.it/portal/node/990932

jeff101's picture
User offline. Last seen 19 hours 21 min ago. Offline
Joined: 04/20/2012
Groups: Go Science

The shared/saved structure viewer could be inside the Foldit interface.
It could look like the spinning 3D *.gif movies of the various amino acids on the page below:
http://swift.cmbi.ru.nl/teach/aainfo/struct.shtml

jeff101's picture
User offline. Last seen 19 hours 21 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Status: Closed » Open
Type: Suggestion » Suggestion

The rotating 3D images could have a standard coordinate system. If the protein has N amino acids in it, the origin could be the center of mass of the protein, the z axis could be from the origin to alpha-carbon #1, and the x axis could be determined so that alpha-carbon #1, alpha-carbon #N, and the origin all lie in the xz plane.

Joined: 07/21/2013
Groups: Beta Folders

I was very impressed with Jeff101's makehelix recipe. In part of it you have the lines:

Assuming CA of m100 is at (x,y,z)=(1.1590344132191,2.3,0).
Assuming CA of r101 is at (x,y,z)=(2.6590344132191,-0.39939080863394,2.2650578319281).
Assuming CA of r102 is at (x,y,z)=(4.1590344132191,-2.1612930278076,-0.78664632964904).

I had loaded some pdb files from the rcsb pdb archive and lost them. I did a file search for pdb files and found several in my Foldit directory. I am running Windows 8 from Microsoft. On my C drive is a Foldit directory with the subdirectory and file.

"C:\Foldit\cmp-resources-ac340afdb1b958531e8251c093261ac5\resources\levels\0000000014.ir_puzzle.pdb"

The contents of this file looks like a PDB format file with the xyz coordinates of the atoms.

ATOM 1 N ASP A 1 -28.097 -1.036 10.434 1.00 0.00
ATOM 2 CA ASP A 1 -26.880 -0.199 10.231 1.00 0.00
ATOM 3 C ASP A 1 -25.993 -0.684 9.091 1.00 0.00
ATOM 4 O ASP A 1 -26.396 -1.531 8.289 1.00 0.00
ATOM 5 CB ASP A 1 -26.051 -0.143 11.516 1.00 0.00
ATOM 6 CG ASP A 1 -25.505 -1.490 11.971 1.00 0.00

The files have a “last modified” time stamp of “7/19/2008 10:53 AM”
And names like “0000000023.ir_puzzle”. I have written a LUA script that reads them. This LUA reads files in DOS mode not part of Foldit. It scans three of these files and three files from the RCSB PDB archive and can not seen any difference. I do not know how to relate these file names to a specific Foldit puzzle or contest that they are part of. I am in the process of checking if any dextro amino acids are created after rebuilding. Trans amino acids were created in the previous version of Foldit, but I have not seen any in the most recent Foldit. I would like to run a LUA script within Foldit and see how the gnarled alpha helices are created when wiggling.

I am split between a wide open Foldit and the comments in the 990932 discussion.

Joined: 09/24/2012
Groups: Go Science

Not sure I understand all consequences of this proposition and debate, but yes, I would find funny to get some rotating 3D preview of shared puzzles, or further on the Wiki for top solution.

Just to reply to brgreening, I wonder if we are not here at the limit of the Foldit etiquette: opening the space for a kind of cheeting (from the game point of view)? it might be good for Science to be able to evolve outside of Foldit, but then we loose most "normal" not specialist players? It turns to Science, not game anymore? I don't know, it's only a feeling.

Joined: 04/15/2012
Groups: Beta Folders

Mt thought is why not have a separate mode where you can use outside tools? Just a thought, but...

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