Direct way to anomalous scatterers

Abstract

Abstract The first step in solving macromolecular crystal structures by multi- or single-wavelength anomalous diffraction methods is the location of the anomalous scatterers. This can be done by direct methods, using either Bijvoet differences within the single data set, or anomalous scattering amplitudes estimated from measurements at several wavelengths. The calculations suggest that Bijvoet differences are equally successful for this purpose as anomalous amplitudes, F A, which theoretically should be more suitable. The calculation of F A values is susceptible to the accumulation of errors contained in the individual intensity measurements at several wavelengths and in the inaccurate estimation of the anomalous atomic scattering corrections, f′ and f″. Direct methods often give better results at resolution lower than the full extent of the diffraction data limit. This may be attributed to the enhanced accuracy of measurements of the strong, low resolution reflections and to more effective phase refinement and propagation through Σ2 relations.