Serine in beta sheets

Case number:699969-993189
Opened by:mbinfield
Opened on:Thursday, July 19, 2012 - 17:01
Last modified:Friday, July 20, 2012 - 01:09

There might be another problem with the scoring function, along the same lines as the issue with design puzzles favoring the large aromatic side chains.

From what I've read, serine is considered a structure-breaking amino acids and does not normally show up in sheets or helices. So, if serine doesn't belong in a beta sheet in nature, I wouldn't expect the mutate function to be substituting it into beta sheets very often. However, I've noticed that in freestyle design I'm often getting serine in my beta sheets. If it were just one or two, that would be realistic, but I'm looking at a ten-segment beta strand with four serines, and in one case recently I had a sheet that was almost entirely serines. That can't be right.

(Thu, 07/19/2012 - 17:01  |  2 comments)

brow42's picture
User offline. Last seen 2 years 1 week ago. Offline
Joined: 09/19/2011
Groups: None

Check this out! Serine self-bonds in a hexagonal trimer barrel. Of course alanine was lower scoring. Still pretty cool coincidence of positioning.

Joined: 12/06/2008
Groups: Contenders

This might be related to the same scoring aberration that made an aromatic-rich 10-residue helix score very highly on puzzle # 571. I'm rating this up a notch, since higher-scoring proteins with unrealistic sidechain combinations won't be helpful in the lab.


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