The principle of single crystal X-ray structure analysis is the same for organic/inorganic materials and proteins. However, although the steps of structural analysis are the same, there are major differences in the method of executing each step between structural analysis of small molecules and proteins. One of the steps where there is a major difference is the phase determination method for solving the phase problem, regarded as the central problem of single crystal X-ray structure analysis.
The main method of phase determination in small molecule structure analysis is the direct method of inferring phase via statistical processing of diffraction intensity. In protein structure analysis, on the other hand, phase determination using the direct method is impractical, and thus phase is determined experimentally. In the direct method, a roughly assigned initial phase is improved by using a phase relation which takes the magnitude of diffraction intensity as a clue. Therefore, in protein crystals, which have a comparatively large lattice, and limited atomic species and deviation of the electron distribution in the crystal, the magnitude of diffraction intensity is small compared with a small molecule crystal, and thus it is difficult in principle to apply the direct method.
As an experimental phase determination method for protein structure analysis, the MIR (Multiple Isomorphous Replacement) method serves as the classical approach. The MIR method derives the phase of the target (native) protein by using the slight shifts in phase which occur when heavy atoms are incorporated. This paper explains the MIR method, and the MAD (Multiple Anomalous Dispersion) method which uses the wavelength dependence of anomalous dispersion.