Closed since almost 12 years ago
Advanced Overall DesignGlycine is an unusual amino acid. Of the 20 standard amino acids, it is the only one that does not have handedness: all others cannot be superimposed on their mirror images, but glycine can. This means that glycine is just as happy forming one structure -- (an ordinary, right-handed alpha-helix, for example) as it is forming the mirror image structure (a backwards, left-handed helix, for example). This extra flexibility and the competition with other, energetically equivalent states makes it hard to lock glycine in one conformation, meaning that glycine tends to disrupt secondary structure.
As a special, extra-hard challenge, we're asking you to try to design a protein that is able to fold, with the restriction that every set of 9 residues has a stretch of 3 glycines in it. This means that regions of secondary structure are likely to be very short. We will be testing top designs with our structure-prediction algorithms. The usual rules apply: we want well-packed, void-free, hydrophobic cores and nice, polar surfaces. Good luck!
It's possible that sheets will work better than helices for this puzzle – I don't know! Be aware, though, that although Foldit will let you put the gly-gly-gly sequence into a helical or sheet conformation, glycine tends to break secondary structures, and should be relegated to loops.
Is there a way of enabling idealize on the glycine segs? I get locked seg messages when I click the (!) thingys.
No remix tool available? D:
I played around with it and noticed the same thing. I think the idealize tool itself still works on those segments, as does clicking on a yellow exclamation mark indicating a cis peptide bond, but yes, the red exclamation marks don't seem to be working in the sequence-locked segments. I'll mention it to the developers tomorrow morning at our Foldit meeting – only the sequence should be locked in those segments.
I wasn't sure whether to include the remix tool for this one or not. My guess is that, based on how it works, it wouldn't necessarily produce great results for the poly-gly sequences, so I left it off. The rebuild tool should still do the trick, though.
