 |
 |
Hi Artem, Thanks for your response, but how do you mean to do the ³normalization² with the other TLS related parameters. I have looked at parameters for each of my TLS groups (3 per monomer) dealing with translation and libration along 3 orthogonal axes and a correlation of the 2- a screw motion as well as calculated the mean values of the tensors for each TLS group by looking at the trace of each tensor. The anisotropic displacement parameters for each atom are calculated by decomposition of these TLS tensors in the first place, such that they are consistent with the motion of the rigid body . To clarify, I did not do individual anisotropic refinement -what I did is refine all the structures using identical protocols in Refmac using restrained refinement including TLS refinement with 3 TLS groups/monomer with a maximum likelihood target, followed by restrained coordinate and residual isotropic temperature factor refinement. Charu On 8/25/08 8:25 PM, "Artem Evdokimov" <[log in to unmask]> wrote: > I would take into account the magnitude and direction of other TLS-related > parameters, so that you can normalize the motion with respect to the overall > motion of the contents of the unit cell. Having done that, I would consider > it entirely valid to compare between crystals. > > I am probably entirely wrong :) > > Artem > > -----Original Message----- > From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of > Chaudhry, Charu (NIH/NICHD) [F] > Sent: Monday, August 25, 2008 10:29 AM > To: [log in to unmask] > Subject: [ccp4bb] TLS refinement > > Hello, > > I have a question about comparing anisotropic displacement parameters for a > series of 6 mutant and a wt structure with diffraction from 1.2-1.5 Å. The > space group and unit cell for the series is the same, and most xtals were > obtained by seeding from a hit for one mutant. There are two mols in the ASU > which form an NCS dimer and the mutations alter dimer contacts in the > xtals, and also alter dimer Kd measured by AUC. However, some of the mutants > do not measurably dimerize in solution, but still form dimers in the xtal > lattice. > > When I study the individual anisotropic displacement parameters (derived > from decomposition of the TLS tensors using TLSANL) of the structural > elements making up the dimer interface, the magnitude and anisotropy of the > displacements are increased for the complexes that form weaker dimers or > don¹t dimerize at all in solution. I interpret this as a reflection of the > weakened or non-existent interaction between monomers observed in solution > for these complexes. But the question is how valid is it really to compare > across-crystals? The total anisotropic displacement parameter U(tot) = > U(cryst) + U(int) + U(atom) +U(TLS) so presumably cross-crystal differences > are being accounted for by the single anisotropic parameter applied to the > unit cell in Ucryst which would suggest that drawing such conclusions across > crystals may be reasonable. Any insights or thoughts would be greatly > appreciated. > > Thanks very much, > Charu >