Filter rejects my green glycine

Case number:671071-2003728
Topic:Game: Other
Opened by:Susume
Opened on:Thursday, April 20, 2017 - 04:12
Last modified:Sunday, April 23, 2017 - 12:42

I have a glycine in a loop which is green in ABEGO coloring. It is being rejected by the SS Design filter, as if foldit thinks it is a sheet or helix. It does have two backbone bonds to the adjacent helix, which maybe makes it seem like a helix or sheet, but it had the same two bonds before I wiggled, and it was not rejected by the filter then. In any case, being green should be enough to make glycine permissible, no matter what bonds it has.

One pic is before wiggle (2 backbone bonds on glycine, filter OK) and one is after wiggle (same bonds, filter rejects glycine). Third pic shows list of permitted AAs, which includes proline (which implies a loop) but not glycine. Rama map shows it is a perfectly good green.

Shared solution with scientists named "1367 s2 invalid green glycine".

foldit_1492660367.png455.44 KB
foldit_1492659543.png457.35 KB
foldit_1492661065.png510.95 KB
(Thu, 04/20/2017 - 04:12  |  2 comments)

bkoep's picture
User offline. Last seen 19 hours 14 min ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff

Thanks for reporting this Susume. Foldit (or rather, the DSSP algorithm) indeed recognizes the glycine as part of the helix. If you use the Auto Structures tool, you'll see that Foldit assigns the residue as a helix.

The hydrogen bond in question (with the hydrogen of the GLY amide) seems to be right on the border of what should, and should not, be considered a "hydrogen bond." The geometry of the bond is very poor, and you'll notice that the Bonding subscore for this residue is negligible.

In general, I agree that a positive φ torsion (green or yellow ABEGO color) should probably exempt a residue from "Helix" status, but that might not be so simply implemented. If this problem continues to crop up regularly, we could probably make a specific exception for this in the SS Design filter.

tokens's picture
User offline. Last seen 2 years 8 weeks ago. Offline
Joined: 11/28/2011

Same issue was reported here:


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