Knowledge of the energy-accessible conformational preferences of molecules and chemical substructures is a vital precursor in molecular modelling, drug discovery, protein-ligand docking and crystal structure prediction. The CCDC has been at the forefront of research that establishes crystal structure conformations as accurate and relevant information for these purposes. We have devised methods for conformational analysis from large volumes of structural information which include torsional plots (both Cartesian and polar), but most importantly refine and apply multivariate numerical. In addition we have done considerable work to establish the relevance and transferability of solid-state crystal conformations to other phases, by comparing them with experimental gas-phase data and with the results of (in vacuo) high-level ab initio calculations. This work has shown that crystal conformations are rarely affected by crystal packing forces, except in a very small number of (now) well-known classes of compounds (e.g. substituted and unsubstituted biphenyls, cyclobutanes, etc).
The conformation of a molecule is of particular importance in the field of Crystal Structure Prediction (CSP) as for large molecules the main difficulty is in finding the correct conformations to investigate rather than in determining the best molecular packing arrangement for a given conformation. The Cambridge Structural Database contains a vast amount of information relating to intramolecular structure and thus can provide vital input into which areas of conformational space are actually feasible for molecules in the solid state. Current and future research in the conformational area includes (a) a continued search for as yet unrecognised crystal conformations that are affected by packing forces, and (b) comparison of theoretical methods with CSD-directed conformational searches for use in CSP. Lastly work is on-going to create a conformer generator based on CSD information to ensure correct geometry.