Hello - I have nearly completed a crystal structure at 2.6 A resolution. There are 3 monomers per asymmetric unit (same protein in different conformations), and I have been using non-crystallographic symmetry during the refinement process (REFMAC5). I have been isotropically refining B-values and using TLS. My problem is that the B-values appear to be too tight as a result of non-crystallographic symmetry. For example, an Asp residue has well-defined density, and atomic B factors of about 40. However, the adjacent lysine, with no density for the side chain has B factors of 42. This does not make sense for them to have nearly identical B values, when there is differing density. This may be a consequence of having a residue with well defined density in chain A and chain B, but not chain C. When plotting residue number vs. B factor, the values seem to be too tightly grouped to make sense. Loosening NCS restraints leads to more variability in B factors, a lower R factor, but higher R-free.
My question is how do I keep NCS, and allow B-factors to vary more so I have more reasonable values? I would appreciate any ideas you have. Thank you in advance for your time and assistance. Ana M. Misic Research Assistant Department of Biomolecular Chemistry Department of Bacteriology University of Wisconsin - Madison 420 Henry Mall, Rm 230 Madison, WI 53706 Lab Phone: (608)265-9282 Fax: (608)262-9865
