Predicting the ‘missing cone’ in protein crystallography – Amplitudes from a structure with high-density solvent

Abstract

Abstract Published 6 Å X-ray diffraction data from rubredoxin were used to base trial phase extensions within a ‘missing cone’ of the reciprocal lattice. The intensities from a protein with a high ammonium sulfate in the solvent space were used as if they had been collected by tomographic sampling within a ± 45° planar tilt restriction from the hk0 net. These amplitudes within the permitted sample space, combined with crystallographic phases from a protein backbone model, were used as a basis for the Sayre equation to predict missing amplitude and phase values. The Sayre expansion with amplitudes (but not phases) partially representing a reverse contrast structure was found to give a reasonable prediction of missing phase terms, with a mean difference of 68.6° for 36 reflections within the data set of 127 reflections, or 49.4° for the 12 most intense reflections. Oddly enough, the overall phase accuracy is better than if ±60° tilt data are used for the extension but the prediction of missing amplitudes is worse for the ±45° experiment.