NAME: Marc Delarue
AFFILIATION: CNRS Institut Pasteur, France
CONTACT: delarue@pasteur.fr
TITLE: "Normal Modes, Elastic Models and Molecular Replacement"
ABSTRACT: Using Normal Modes for crystallographic refinement is a relatively old idea, but, curiously, it was limited up to now to the refinement of B-factors (see Go and coll, J. Mol. Biol., 1992, 225:457 and 477).
We have devised and implemented (Delarue and Dumas, 2004, PNAS, 101:6957) a crystallographic program that refines the amplitudes of a handful of the lowest frequency normal modes of an initial model against diffraction data, using standard conjugate gradient techniques and normal modes calculated from the Elastic Network Model (Tirion, PRL, 1996).
This has the effect to favour collective movements of large amplitudes during refinement, capturing large structural rearrangements such as hinge movements and also shear movements (see Krebs et al., 2002, Proteins, 48:682). As ENM preserves interatomic distances, the model conserves protein-like (secondary structures) features during refinement.
This program generalizes the classical rigid-body refinement and has a much larger radius of convergence, while still using a small number of degrees of freedom (typically 6+10). The first 6 normal modes with zero frequency can be used as the 3+3 translational and rotational degrees of freedom of rigid-body refinement.
The program can be used both with structure factors amplitudes (X-Ray) or with phased structures factors (cryo-EM). A web site (http://lorentz.dynstr.pasteur.fr) has been implemented (Nucl. Ac. Res., 2006, 34:W52) to allow for
1) calculating and visualising ENM Normal Modes by Pymol for any PDB file
2) generate a series of decoys with variable amplitudes along the lowest frequency normal modes
3) perform refinement against X-Ray or cryo-EM data (NOMAD-Ref).
The program can be used routinely to identify the true solution among a list a potential solutions of MR, in the spirit of PC-Refinement (Brunger, 1995, Acta Cryst D, 51:740).
Other similar web-sites applying the same ideas in a slightly different implementation will be presented and compared, such as
- NORMA by K. Suhre, J. Navaza and Y.H. Sanejouand (Acta Cryst D, 2006, 62:1098)
- ElNemo by K. Suhre and Y.H. Sanejouand (Nucl. Ac. Res, 2004, 32:W610 and Acta Cryst D, 2004, 60:796)
- NMFF by F. Tama and C. Brooks (J. Struct. Biol., 2004, 147:315)
- a program by K. Hinsen and coll. (Biophys. J., 2005, 88:818).