Optimization of selenium substructures as obtained from SHELXD.

Using the signal of naturally inbuilt or artificially introduced anomalous scatterers to derive initial phases in a macromolecular crystal structure determination has become routine in recent years. In the context of high-throughput crystallography in particular, MAD and SAD (multiple- and single-wavelength anomalous dispersion) methods are central tools. For both techniques, a crucial step is the determination of the substructure of anomalous scatterers; subsequent phasing procedures will profit from a substructure model that is as accurate as possible. The choice of the subset of the diffraction data to be used for the substructure determination has a strong influence on the quality of the substructure and can make the difference between success and failure. The accuracy of selenium substructures obtained using F(A) values or various anomalous differences truncated to different resolutions has been investigated by comparing the sites determined by SHELXD with the selenium positions in the refined models. Based on the analysis, some recommendations for obtaining accurate and precise substructures are derived.