Do walkers stiffen the protein? Written by Krulon
I'm not sure I agree with the concept of stiffening. So I'll throw out some thoughts and let others comment.
I don't think wiggling "stiffens" a fold. If it did, then wiggling a denovo would make it into a stick you could not bend. Rebuilding seems to happen pretty easily, pulling segments seems to be pretty easy also.
I think most people use the word stiffen when they have folded a protein into
a "local minima", which is a position from which it is difficult or almost
impossible, to obtain anymore points, unless you pull it all apart and make
major changes. Theoretically each protein can have hundreds of positions
which could be considered local minimas. Loop rebuilders, banders,
wiggling, just won't find anymore points. Changing the behavior clash
setting may let a sidechain "slip" past another sidechain, allowing the
protein to settle into a higher scoring position.
So, it does not bother me to use a settler or walker early on. I don't
think it makes a protein stiff. I can still pull it apart or rebuild
segments easily… However… settling does take time and eventually you are
going to rebuild parts of the protein, so I think, settling a protein early
and often, is a waste of time, unless you have pulled segments together with
bands and you are trying to settle it so that your protein will stay in the
shape you want.
As an evolver I like to first settle a shared protein to evolve it. Then I know, I will get some credit for my effort. After it is evolved then, I
try the longer running overnight scripts. I don't like to run a script all
night just find it did not evolve at all.
It is also nice to be able to settle a protein in the last hour to get some points to keep up with the people who are passing you. On 433, sweeper
gave me 23 points and took me from #17 to #10 in the last hour.
However, sometimes when rebuilders or banders are not working for me, I
have been able to settle and then sometimes the banders and rebuilders will
begin to pickup additional points afterwards using the same rebuilders or
Seagat2011 has a public shared recipe "ST-Softwalker" which he says will "gently LWS (Local Wiggle Segments) protein in a way which prevents stiffening." So I guess he believes in "stiffening". All his recipe appears to do, is local wiggles, global wiggles, and shakes, so I don't think there is any anti-stiffening magic in there other than maybe once the wiggle gains stop, he restores the protein to an earlier position which still had some wiggle left in it. (I did not analyze his script completely.)
More techniques at > The Foldit Labs