I'm enjoying these little design puzzles. I think 3 residues is a good size for the binding probe. Would it be possible to leave the protein backbone frozen, but to unfreeze the sidechains near the binding site? it's annoying to be just out of range of some great big flimsy lysine.
The small ligand wiggles slowly….I've literally watched it go up by 0.001 point per second for 90 minutes. Looks like jumping in with a local wiggle speeds that up though, but I'll have to modify scripts to do a local wiggle instead of a global wiggle. I miss the wiggle slider (sounds of teeth gnashing as people read this).
Anybody else have any comments?
I agree with your comment about the sidechains. I would be surprised if these sidechains did not react to the intruder coming in… and that makes a difference, I think, in the way the two interact.
First of all good to hear you like the small design puzzles :) I can definitely imagine it can sometimes be a pain to see the backbone is completely frozen. In the CTLA-4 puzzle (539) we based this decision on several pieces of information. First of all the available crystal structures of bound and unbound proteins; information of this structures provides clues for the "floppiness" of particular residues or the whole structure. In this particular case (539) it guided us into a more rigid structure. However as bertro points out, it would be unlikely that these residues would not interact with a binding partner. So we try to balance this out in the right way. Furthermore because this is the first puzzle in the series we did not, for a start, want to introduce too much variability in the puzzle.
Thanks for this suggestion!
