Crash on 1922 pulling on a cysteine

Case number:845829-2010883
Opened by:NinjaGreg
Opened on:Saturday, November 28, 2020 - 03:03
Last modified:Saturday, December 5, 2020 - 20:13

had banded two cysteine to try to form the bridge, and did the "click and drag" to pull one, and it crashed. Files attached.

(Sat, 11/28/2020 - 03:03  |  3 comments)

NinjaGreg's picture
User offline. Last seen 3 days 5 hours ago. Offline
Joined: 05/21/2010
Groups: Go Science

The files.

Joined: 09/29/2016
Groups: Gargleblasters

Greg, was the Band you mentioned, by chance attached to the "Bondable H" (white tip) of the CYS?

If so, that's one of the two known crashes I documented awhile back that affect CYS and Disulfide Bonds. (It's in the Bug Doc I keep for the Devs; if you need/want the link again, feel free to PM me here or on Discord, as it's not a public facing doc for various reasons.)

To quickly convey for everyone what exactly those two instances are and what is needed to encounter (or avoid!) them:

  a) Requires "Show Bondable H" or "Show All H" enabled in View Options
  b) Requires a Band to be placed between the White Tips of CYS. [It may only require a Band on just ONE of the CYS, but I've not tested to confirm.]
    - Note: Also, I think the band needs to be Enabled.
  c) Create the Disulfide Bond (BRidge), however you so choose. Be it pulling the sidechain, pulling the protein, moving a cut portion of the protein, shake, wiggle, wiggle sidechains -- in other words, anything that creates the bond will be sufficient!
  d) Client Crash! Due to someone else's discovery some months back, this happens because the White Tip suddenly gets removed when two CYS bond, and since there was an active band placed there, Foldit tries to continue calculating its impact, but it's no longer there; thus, the crash.
NOTE 1: This only happens when banding specifically to the White Tip! If you band between the yellow tip (H not visible) or the yellow/white elbow (H enabled), you are SAFE and will not crash!
NOTE 2: Admittedly, I haven't actually checked this in a few months, and the more recently fix for the Bands "Out of Bounds" crash relating to recipes, MAY have fixed this one...

  a) Requires a Cysteine which is already Bonded/Bridged.
  b) Pretty sure it does not matter if "Show Bondable H" is being used.
  c) Use the "Pick Sidechain" tool on a Bonded Cysteine to bring up the Rotamer Tool
  d) Select a different Rotamer choice from the list. REQUIRED: It has to be a choice that is something other than what it currently was.
  e) Close the Rotamer tool window. (Either click the [X} or click protein background and press Spacebar)
  f) Client Crashes!
NOTE 1: This occurs only when confirming the new selection. You can still open the Rotamer tool and click on the options, but it is only represented locally (the pose change is only graphical, not 'actual') which can spare you the crash if you either just want to look at where it can be moved to, or accidentally open it and catch yourself before making a change (which happens to me often enough).
NOTE 2: If you want to actually use the tool on a bonded CYS, first manually move one of them to break their bond. THEN you can use it on one/both and make the selection you want. The FORMATION of a bond from using the tool will not cause the crash. (go figure!)

TL;DR - If you're using a Disulfide Bridge recipe and Show Bondable H, do NOT let it place bands on the Hydrogen!
If you have a Bond between two CYS, do NOT use Pick Sidechain tool on it to change its Rotamer!
Hope that helps some folks.

NinjaGreg's picture
User offline. Last seen 3 days 5 hours ago. Offline
Joined: 05/21/2010
Groups: Go Science

I don't recall if I had banded to the hydrogens, but it's entirely possible. Let's call it the same bug. Thanks for the explanation!


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