Disulfide bonds are very hard to create at CI 1

Case number:845813-2004907
Topic:Game: Tools
Opened by:Timo van der Laan
Opened on:Tuesday, March 20, 2018 - 21:09
Last modified:Thursday, March 22, 2018 - 19:54

The standard tool I used to do this was bridge wiggle. But when running at CI 1 with the CYS only 2.6 apart bridge wiggle no longer manages to close the bridge. Also connecting the tips with a band strength 10 did not managed to do that.
My two cents: maybe the new scoring function makes a too great class to overcome.
Also pulling by hand never made the bridge. The only way I found was using cuts and move.
Others have managed to make bridge but they where working on low CI.
Can the scoring function be changed when clashing between the tips of CYS is concerced?

(Tue, 03/20/2018 - 21:09  |  5 comments)

frood66's picture
User offline. Last seen 16 weeks 1 day ago. Offline
Joined: 09/20/2011
Groups: Marvin's bunch

I have had no probs at all - all works about the same with bridge creation - MAC user

Joined: 09/21/2011
Groups: Void Crushers

One big difference as you said working on a low CI.

brow42's picture
User offline. Last seen 3 weeks 5 days ago. Offline
Joined: 09/19/2011
Groups: None

I started the new toxin and ran with the default guesses (which are not good guesses but anyways). It wouldn't close all 4 pairs. This doesn't seem very unusual. I don't think it's clashing, but I think it's backbone stiffness and/or the wiggle function hitting its minimum at low wiggle power (it just did nothing instead of building the bond.) I had no trouble making bonds with hand nudges. You could be right, though, if the a new potential function has created a shallow barrier that wiggle won't push through.

If the disulfide routine has a hard cutoff, perhaps that cutoff could be extended slightly?

There are some tweaks I can add to bridge wiggle. I could make the SS band a little shorter, maybe a user scaling. Also, I just did some research and I see that the C-C-S-S angle is typically -60 degrees, and I assume 90 in my band lengths. And, it looks like the bands aren't putting things in the right place because everything is spring-y...the band lengths are for the bonded configuration and the bands seem to pull some central point on each sidechain closer together instead of pulling the sulfurs closer together.

Joined: 09/24/2012
Groups: Go Science

Same observation with bridge wiggle after the latest updates.

I observed that the default band length if Foldit is now 3.5, which seems to be not enough to make a bridge. When I change band length to 0, bridges are occuring more easilly. The distance between Cys ends for bridges is actually about 2.02 A.

Could it be the part of the new problem?

brow42's picture
User offline. Last seen 3 weeks 5 days ago. Offline
Joined: 09/19/2011
Groups: None

The default band length has always been 3.5 A, which is about the C-C distance along an extended backbone, and also about the C-C distance between two sheets. This makes sense, since the main use of bands is to pull sheets together. For a disulfide bond, you need to manually set it to 2.02 to 2.05 (or a bit less to pull it together).


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, Boehringer Ingelheim, RosettaCommons