Knowledge-Based Optimization of Molecular Geometries Using Crystal Structures.

This paper describes a novel way to use the structural information contained in the Cambridge Structural Database (CSD) to drive geometry optimization of organic molecules. We describe how CSD structural information is transformed into objective functions for gradient-based optimization to provide good quality geometries for a large variety of organic molecules. Performance is assessed by minimizing different sets of organic molecules reporting RMSD movements for bond lengths, valence angles, torsion angles, and heavy atom positions.

Deducing chemical structure from crystallographically determined atomic coordinates.

An improved algorithm has been developed for assigning chemical structures to incoming entries to the Cambridge Structural Database, using only the information available in the deposited CIF. Steps in the algorithm include detection of bonds, selection of polymer unit, resolution of disorder, and assignment of bond types and formal charges. The chief difficulty is posed by the large number of metallo-organic crystal structures that must be processed, given our aspiration that assigned chemical structures should accurately reflect properties such as the oxidation states of metals and redox-active ligands, metal coordination numbers and hapticities, and the aromaticity or otherwise of metal ligands. Other complications arise from disorder, especially when it is symmetry imposed or modelled with the SQUEEZE algorithm. Each assigned structure is accompanied by an estimate of reliability and, where necessary, diagnostic information indicating probable points of error. Although the algorithm was written to aid building of the Cambridge Structural Database, it has the potential to develop into a general-purpose tool for adding chemical information to newly determined crystal structures.

Data mining of iron(II) and iron(III) bond-valence parameters, and their relevance for macromolecular crystallography.

The bond-valence model is a reliable way to validate assumed oxidation states based on structural data. It has successfully been employed for analyzing metal-binding sites in macromolecule structures. However, inconsistent results for heme-based structures suggest that some widely used bond-valence R0 parameters may need to be adjusted in certain cases. Given the large number of experimental crystal structures gathered since these initial parameters were determined and the similarity of binding sites in organic compounds and macromolecules, the Cambridge Structural Database (CSD) is a valuable resource for refining metal-organic bond-valence parameters. R0 bond-valence parameters for iron(II), iron(III) and other metals have been optimized based on an automated processing of all CSD crystal structures. Almost all R0 bond-valence parameters were reproduced, except for iron-nitrogen bonds, for which distinct R0 parameters were defined for two observed subpopulations, corresponding to low-spin and high-spin states, of iron in both oxidation states. The significance of this data-driven method for parameter discovery, and how the spin state affects the interpretation of heme-containing proteins and iron-binding sites in macromolecular structures, are discussed.

Generation of crystal structures using known crystal structures as analogues.

This analysis attempts to answer the question of whether similar molecules crystallize in a similar manner. An analysis of structures in the Cambridge Structural Database shows that the answer is yes - sometimes they do, particularly for single-component structures. However, one does need to define what we mean by similar in both cases. Building on this observation we then demonstrate how this correlation between shape similarity and packing similarity can be used to generate potential lattices for molecules with no known crystal structure. Simple intermolecular interaction potentials can be used to minimize these potential lattices. Finally we discuss the many limitations of this approach.