Phase determination in the x-ray diffraction patterns of complex crystals and its application to protein structure.

Abstract

THE phases, or signs in cases of central symmetry, of X-ray reflexions from organic crystals can only be determined unambiguously and without prior assumptions by the use of methods employing heavy atoms or isomorphous replacements. In cases where these methods cannot be used, they have been determined hitherto largely either by means of conjectural molecular models with subsequent refinement, or by the use of various inequalities such as those of Harker and Kasper1 and later elaborations, all of which depend on the existence of discrete scattering maxima corresponding to a finite number of atoms. In very complex crystals such as proteins these methods are inapplicable or of very doubtful validity, as the individual atoms are very numerous–sually more than a thousand–and not easily resolved. Harker has attempted to overcome this difficulty by the use of hypothetical, large, spherical groups or ‘globs’, but these are difficult to justify on physical and chemical grounds.