-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Dear Herman,
you could apply random shifts to the occupancies, coordinates, and B-values and refine a large number of times, e.g. with 50 variations. Then you can calculate the mean occupancies with standard deviations. That's give you an idea about the precision of the occupancies. I am not sure how to get to the accuracy - you would have to figure out how much all the parameters are interrelated, which I dare say is quite impossible - it is easier for the referee to learn what a crystallographic model represents. Best, Tim On 11/22/2013 10:07 AM, herman.schreu...@sanofi.com wrote: > Dear bulletin board members, > > A referee came back with a question about the accuracy of refined > group-occupancies. In the manuscript we describe 3 crystal > structures with resolutions between 2.1 and 2.4 ￅ. In all three > cases, the inhibitor has been fitted in two alternative > conformations and the group occupancy of each conformation has been > refined and the relative occupancies vary depending on the > inhibitor. > > The accuracy will depend on the amount of overlap between the > inhibitor densities and the amount of unique density belonging only > to one conformation and the amount of additional disorder of the > inhibitors, since most of the interactions are non-specific > hydrophobic interactions. Also partially occupied solvent molecules > could lead to errors in the refined occupancies. My gut feeling is > that the error will be somewhere between 20 and 30% or even higher, > but I would be very happy if someone could come with a more > scientific way, however crude, to get an estimate of this error. > > Best regards, Herman > > - -- - -- Dr Tim Gruene Institut fuer anorganische Chemie Tammannstr. 4 D-37077 Goettingen GPG Key ID = A46BEE1A -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Icedove - http://www.enigmail.net/ iD8DBQFSjyOsUxlJ7aRr7hoRAhkNAJ45iZLSy+YQnNeu9PMAStScQbu5MgCgwB/G kH9IVkjsDyFpiEItjMJn8MA= =lSi6 -----END PGP SIGNATURE-----
