> areas of the protein that undergo movement and show general trends. There is > also a server at Yale (http://www.molmovdb.org/molmovdb/ ) that will perform > simulated annealing while morphing between two structures, thereby producing > intermediate structures that violate fewer protein restraints.
You can also just download the script that does this. If you know enough about CNS, this also allows you to change which energy terms are used. I've found that eliminating dihedrals from the equation sometimes produces smoother movies - for example, RNA base flips aren't so sudden - but still more realistic than linear interpolation. This doesn't actually do simulated annealing, only energy minimization. I'd like to figure out how to do generalized targeted molecular dynamics (see for example http://sigler.bioch.bcm.tmc.edu/MaLab/gallery.php) in the context of the morph server, but this is probably a fairly difficult project.
