*** For details on how to be removed from this list visit the *** *** CCP4 home page http://www.ccp4.ac.uk ***
Patrick G. Shaw wrote:
Hello, I have been working for some time on finding a molecular replacement solution for an IGG receptor protein with little to show for it. My search model is a previously published homologue of the protein that my colleagues and I now suspect differs greatly with our crystallized protein about a central hinge angle, despite 97% sequence homology between the two. Any attempts at molecular replacement with the two domains treated separately result in a pdb with the terminal ends pushed together instead of the ends that where severed about the hinge region. This is probably due to the fact that the crystal is very tightly packed and in high symmetry (P 43 21 2).
Dear Patrick, If these solutions look good otherwise, I would see this a problem of inappropriate choice of the asymmetric unit (not your choice, but the one MR gave you). You correctly located two bodies in the asymmetric unit, but the asymmetric unit you end up with has the C-terminal domain of one molecule and the N-terminal domain of another. You want to get an asymmetric unit with N-term and C-term domains of the same molecule and eventually link them together in a single chain. This is easy to fix using O. Load your current inappropriate AU contents into say molecule M. do "sym-set m" and give the cellparam and spacegroup, if O hasn't already deduced them from the CRYST1 record. "center-atom " on one of the severed ends "sym-sph m ;;" or if nothing is found "sym-sph m ; 20" or so until you find the sym mate sym1. The macro will draw all the atoms within your selected radius (20A here). But in actuality O has generated an entire new asymmetric unit in mol sym1. save it: s-a-o sym1.pdb sym1 ;;;;; and use a text editor to combine the domains with the severed regions abutting to generate an appropriate AU. Hope that helps, and that I didn't completely misunderstand the problem, Ed
