Abstract
Complete and highly redundant data sets were collected at different wavelengths between 0.80 and 2.65 A for a total of ten different protein and DNA model systems. The magnitude of the anomalous signal-to-noise ratio as assessed by the quotient R(anom)/R(r.i.m.) was found to be influenced by the data-collection wavelength and the nature of the anomalously scattering substructure. By utilizing simple empirical correlations, for instance between the estimated deltaF/F and the expected R(anom) or the data-collection wavelength and the expected R(r.i.m.), the wavelength at which the highest anomalous signal-to-noise ratio can be expected could be estimated even before the experiment. Almost independent of the nature of the anomalously scattering substructure and provided that no elemental X-ray absorption edge is nearby, this optimal wavelength is 2.1 A.