Incorporation of Partial Structure Information in Macromolecular Crystallographic Phasing

Solving the phase problem can be a limiting step in the structure determination of macromolecular X-ray crystallography. Recent development of the iterative projection algorithm shows that it is a promising tool for phase retrieval. Two approaches for enhancing the iterative projection algorithm are proposed in this thesis for solving the macromolecular crystallographic phase problem. When using a homologous molecule in the molecular replacement method to determine a new macromolecular structure, even if a proper template structure is positioned, the insufficient similarity between the template and target structures may cause its resistance to the conventional molecular replacement method. A modified phasing method is presented with better convergence properties and can effectively overcome the difference between the template and target structures. This is illustrated by the reconstruction of one trial structure, which has also been solved and analyzed in several subsequent molecular replacement methods studies. Uniqueness of the phase problem in macromolecular crystallography limits the ab initio phasing. To enlarge the the radius of convergence of the iterative projection algorithm in macromolecular crystallography, an enhanced density modification method is proposed by incorporating partial structure information. Four trial calculations show a variety of potential applications of this method in novel structure determination and model validation.