Probing conformation and conformational change in proteins is optimally undertaken in relative mode.

Hydrodynamic bead modelling has been widely used in attempts to assess the 3D conformation of proteins in solution. Initially, simple models employing only a small number of beads were used, with a considerable degree of success. Latterly, high-resolution bead models based upon atomic coordinates have been developed, and much more sophisticated questions can in principle be addressed. A detailed analysis is presented of the errors involved in the generation of such models and associated prediction of (translational friction) parameters, and in the practical measurement of these parameters for comparison. It is shown that in most cases, for a particle of only moderate asymmetry, the errors are such that it is not feasible to determine, on an absolute basis, which of a range of candidate conformers is the "correct" one. However, when the properties of the candidate conformers can be compared in relation to those of a "paradigm conformer", whose structure in solution, on the basis of external evidence, can be accepted as correct, then errors cancel and very precise comparisons become possible. The generation of 3D bead models (and hence 3D data files) for a range of candidate conformers is a simple matter, using the existing program MacBEADS, further facilitated by a 3D display module (pro Fit).