Crystal structure prediction molecular mechanics

Since many critical physical and mechanical properties of pharmaceutical compounds are in large part dependent on crystal form, accurate prediction of crystal structure would be a highly valued tool in the pharmaceutical industry. Unfortunately, to date, reliable crystal structure prediction is only feasible for rigid, low molecular weight molecules, which do not represent the size and flexibility of pharmaceutical molecules. 

Two independent determinations of the structure of (-)-epicatechin in the crystalline state have been performed (8,18). The reports describe the same basic structure, but that of Fronczek et al. (8) provides higher resolution. The heterocyclic ring is approximately a half-chair with C-2 26.3 pm above and C-3 49.5 pm below the mean plane of the fused ring system (8). The dihydroxyphenyl substituent at C-2 adopts a pseudoequatorial position. Thus far, (+ )-catechin has resisted crystallization into a form amenable to a structure determination. Molecular mechanics calculations that reproduce the observed structure of (-)-epicatechin predict that the heterocyclic ring in (+ )-catechin has a similar conformation (9). 

Based on the same underlying principles as the molecular-based quantum methods, solid state quantum mechanics represents bulk material using periodic boundary conditions. The imposition of these boundary conditions means that it becomes possible to expand the electron density in periodic functions such as plane waves, as an alternative to the atom-based functions used in the molecular case. The efficiency of the calculations is enhanced by the use of pseudo-potentials to represent the core electrons and to make the electron density as smooth as possible near the nucleus, hence reducing the complexity of the plane wave expansion of the electron density. Because of the number of choices available for pseudo potentials, basis sets and whether calculations are done in real or reciprocal space, there are many choices of software for performing solid state quantum mechanical calculations. A few examples which have been used in crystal structure prediction include the Vienna ab initio Simulation Package (VASP), CASTEP and CRYSTAL. " A wider ranging introduction to the area can be found in the references. ... 

The subject of this study is particularly challenging for any crystal structure prediction methodology for several reasons. First of all the molecular entity in the solid state (a zwitterion) is different from that in the gas phase. This means that the transfer of molecular information from quantum mechanical... 

Molecular mechanics force fields have largely been parameterised using the best available data from the gas phase and (in some cases) from liquid phase or solution data. The question therefore arises as to how applicable molecular mechanics force fields are to predicting structures of molecules in the liquid crystal phase. There is now good evidence from NMR measurements that the structure of liquid crystal molecules change depending on the nature of their... 

Despite the vast number of reported syntheses, crystal structures, properties and applications of monomeric Pcs, it is still difficult to clarify clearly the synthetic mechanisms of Pcs in various conditions, to predict fully the supramolecular structures of Pcs in the solid state (except for a very few types of simple Pcs), and to elucidate completely the correlation between molecular structure and properties. The large-scale preparation and separation of some novel Pcs with interestingly properties is still very hard. On the basis of the great potential applications of Pcs in high-tech fields, exploitation of multi-functional Pc materials needs to be strengthened in the future. There is still plenty of room for further investigation of Pc chemistry. 

Current efforts toward oPE foldamers have focused on developing a computational model that will allow useful predictions of structure. This model was validated by comparing computational results to previously published experimental results [101]. For both crystal structure data and methyl- and hydrogen-series mPEs [25,57], the MMFF molecular mechanics system [102,103]... 

The theoretician uses these programs to predict structure, either of single molecules or of assemblages of molecules, using X-ray or NMR data, when available, to test his predictions (13-15). It has been known for a long time that even the earlier molecular mechanics programs can predict the structures of certain types of molecules with excellent reliability. For the cyclic alkanes, an accuracy comparable to that of the best X-ray crystal structure analysis can be obtained. In fact, the method is more widely applicable since neither compound nor crystals are necessary (1 . 

There are two apparent exceptions to the rule of Hooker et al. 1,170 These are c(-Ala-Ala-) and c[-(Me)Ala-(Me)Ala-]. The X-ray structure of c(-Ala-Ala-) 171 172 gives a positive p, but the NV couplet is observed to be negative. However, molecular mechanics calculations 173174 predict a negative p for c(-Ala-Ala-) and its A-methyl analogue, in agreement with the CD assignment, so crystal packing effects are responsible for the apparent exception. 

---