I agree with Ethan. In philosophy, NMA is a useful analysis to study low frequency collective motions. That is true by taking a stand-alone structure and explore such motions of biological interest. Domain motions in the crystallographic environment need not necessarily correspond to those of the isolated molecule. Also, please remember, many a times the most significant modes (say mode 7 or mode 8) in NMA do not represent the motions of biological interest. One is usually recommended to analyze many significant modes to extract the useful information, and in the extremes of argument, making at times this as a subjective exercise.
In any case, I would strongly argue for TLSMD, as Ethan has pointed out, as that represents modeling disorder in crystallographic environment. In one of the examples where we were studying the relationship between NMA (for a stand-alone molecule) and TLS, we could correlate the two for the 11th and the 12th mode, suggesting clearly that 7th- 10th modes did not show a good correlation. By incorporating the 7-10th modes in crystallographic refinement, we would have clearly not made the model better! Shekhar On Tue, Oct 21, 2014 at 6:22 AM, Ethan A Merritt <merr...@u.washington.edu> wrote: > On Monday, 20 October, 2014 18:10:03 Appu kumar wrote: > > Dear CCP4 Users, > > I seek your valuable advice and suggestion in carrying out the normal > mode > > structure refinement which manifest the dynamics of protein as linear > > combination of harmonic modes, used to describe the motion of protein > > structure in collective fashion. Studies suggest that it is highly useful > > in refining the protein structure which harbors a considerable magnitude > of > > flexibility in atomic position owing to high thermal factors. > > Therefor I want to know is there any software/script available to execute > > the normal mode of refinement. Thanks a lot in advance for your > imperative > > suggestions > > The previously published examples of normal-mode refinement that I know > about used private external programs to generate thermal ellipsoids for > each > atom, and then used those as fixed ADPs while refining coordinates in > refmac or similar standard program. Again speaking only of the examples > I have looked at in detail, the result was "better" (had lower R factors) > than a conventional isotropic refinement but was not nearly as good as a > multi-group TLS refinement of the same structure (TLSMD + refmac). > > On the other hand, there is a quite different way normal modes can be used > in refinement. As I understand it (perhaps Garib will add addtional > details) > the "jellybody" refinement mode of recent refmac versions can be viewed > as restraining the model shifts to be consistent with the principle normal > mode. > In this way the normal mode contributes to the path of the refinement, > but is not explicitly part of the final model. > > So it may be that using TLSMD + refmac jellybody TLS refinement > would get you the best of both approaches, though I have not gone back > to look again at the published example structures since the advent of > jellybody refinement. But note that jellybody is primarily useful when > you already have a high-qualityl, good geometry, starting model. > > Ethan > > -- > Ethan A Merritt > Biomolecular Structure Center, K-428 Health Sciences Bldg > MS 357742, University of Washington, Seattle 98195-7742 > -- Shekhar C. Mande (शेखर चिं मांडे) Director, National Centre for Cell Science Ganeshkhind, Pune 411 007 Email: shek...@nccs.res.in, direc...@nccs.res.in Phone: +91-20-25708121 Fax:+91-20-25692259
