On Saturday 18 October 2008, Leoanrd M Thomas wrote: > To clarify a bit, the protein is not cyclic, it is two of the wild type > protein molecules joined together by a short peptide linker creating a > "new" protein. > > Here is my attempt at an ascii sketch, > > 2-fold 2-fold > __ > ( ) ( ) Linker > ( ) ( ) > ( ) ( ) > ( ) ( ) > ( ) ( ) > Linker (__) ( ) > > When looking at the crystal imagine half the unit cells of the crystal > have the linker on the "bottom" of the protein and the other half have it > on the top.
Your description seems to contain the answer to your question. Since half of the time the linker is in one place, and half of the time it is in the other place, then that is what you should describe and refine. Place the linker in both places in the model, but at half occupancy. The fact that the linker spans two asymmetric units is not relevant. We have refined systems that were very similar to your case, where bi-functional ligands bound with one end on one protein molecule and the other end on a different protein molecule, spanning two asymmetric units. We had the further complication that the specific ligand binding sites on the two multimeric protein molecules were not crystallographically equivalent. > The modified protein crystallizes in the same space group and > unit cell as the wild type. > > I am just not sure how to describe it or for that matter refine the > system. One half of the new molecule is in the ASU while the other half > is related by the crystallographic two fold. > > We have run the usual checks for twining and making sure we are in the > proper space group. We would not have this problem if we ignored the > higher symmetry, of course it would be wrong. It refines nicely in both > the space groups but the overall processing statistics really indicate > that it is in the higher space group. > > Cheers, > Len > > > > Leonard M. Thomas Ph.D. > Director, Macromolecular Crystallography Laboratory > Howard Hughes Medical Institute > California Institute of Technology > Division of Biology > 1200 E. California Blvd. MC 114-96 > Pasadena, CA 91125 > 626-395-2453 > [EMAIL PROTECTED] > http://www.br.caltech.edu/cmclab > -- Ethan A Merritt Biomolecular Structure Center University of Washington, Seattle 98195-7742
