If you do not want to get deeper into the technology and all you need from time to time is a list of all relevant crystal packings ranked according to their lattice energy, then we can run all calculations for you on our own 1536 core Intel Xeon cluster.
After having signed a confidentiality agreement, we take a look at your molecule and provide you with an estimate of how much time it will take to carry out a complete in silico screen in all 230 space groups with one molecule per asymmetric unit. If you accept, we run the calculations on a cluster that is completely disconnected from the outside world. As deliverables, we provide a report, a list of all crystal structures predicted within a certain energy window and a tailor-made force field.
Based on the performance of our code during the screen with one molecule per asymmetric unit, we provide you with a quote for a screen with two molecules per asymmetric unit. Screening with two molecules per asymmetric unit is significantly more time consuming that screening with one molecule per asymmetric unit. The feasibility depends on two factors: the accuracy of the tailor-made force field and the number of different crystal packings found in a given energy window. Neither of these pieces of information is available before the screen with one molecule per asymmetric unit has been completed.
A screen with only one molecule per asymmetric unit can already produce valuable results: 88.5% of all crystal structures consisting of a single molecular species and having no molecule on a special position crystallize with one molecule per asymmetric unit. 99.8% of these crystal structures contain either one or two molecules per asymmetric unit.
The figure below shows three molecules approved by the FDA between 2004 and 2008 with estimated calculation times at full load for one molecule per asymmetric unit.
Click here for more estimates.
Assuming that you have already determined at least one crystal structure of your compound, an easy check is available to you to assess the reliability of our calculations: that experimental crystal structure must be present among the first few most stable predicted crystal structures. For us, every contract research study is a blind test .
We will first send you the list of predicted structures and in return you may send back your experimental data. The comparison of the predicted structures with the experimental data is included in the final report. If applicable, the final report will also suggest experiments to generate additional crystal forms and contain a risk assessment for experimental appearance of a more stable crystal form.
Next page: Software