CCDC Blind Test Showcases Major Advance in Crystal Structure Prediction Methods
November 3, 2015
The Cambridge Crystallographic Data Centre (CCDC) announces that the results of its 6th blind test of crystal structure prediction methods demonstrate significant advancement in crystal structure prediction methods in comparison with previous tests. This year, structures of all of the test systems, which included the complexities of polymorphs, salts and hydrates, were generated by one or more methods. In addition, a number of the target experimental structures were predicted to be the most stable form.
Participants in the blind test were invited to predict the full three-dimensional crystal structures of five previously unpublished compounds, starting solely from the chemical diagram and basic crystallization conditions. The participants came from a record number of 25 different groups, spread over 14 countries (Argentina, Austria, Canada, China, Germany, India, Italy, Japan, The Netherlands, Poland, Russia, UAE, UK, and USA).
The subjects of the test included a co-crystal, a large flexible molecule, a salt hydrate, and a drug molecule with five known but unpublished polymorphs; the drug molecule was released by a major pharmaceutical company especially for the blind test. The inclusion of the polymorphic drug molecule for the first time highlighted the value of crystal structure prediction-derived knowledge for identifying polymorph issues and opportunities for molecules of commercial importance. The salt hydrate is the first three-component system in the blind test, and potentially one of the hardest to predict, to date.
“Overall the target structures in this blind test were significantly harder than for previous tests,” commented Colin Groom, Executive Director of the CCDC. “So the results this year are especially impressive. The increased participation and breadth of methods adopted underline the importance of crystal structure prediction methods, in particular for pharmaceutical R&D.”
Professor Richard Cooper, Oxford University, who selected the target systems for the blind test added, “These targets were not model systems but were selected to represent real-life challenges. The remarkable results of this blind test demonstrate the huge potential of crystal structure prediction methods for informing experimental solid-state chemistry.”
A paper detailing the full results of the blind test will be published as part of a special issue on organic crystal structure prediction of the International Union of Crystallography journal, Acta Cryst. B.
3D structure for Molecule XXIV: a chloride salt hydrate of diaminomethylthio acrylic acid