MAKING THE MOST OF YOUR IN-HOUSE DATA
Peter Müller, G.M. Sheldrick
Institut für Anorganische Chemie, Universität Göttingen, Tammanstr. 4, D-37077 Göttingen, Germany; peterm@shelx.uni-ac.gwdg.de
Keywords: Ab initio, Data Collection, Redundancy, Structure Quality, HEW-Lysozyme.
This project started as a teaching exercise, but turned out to be more interesting. In order to find out how to get the best possible dataset out of protein crystals with standard laboratory equipment, we have collected several highly redundant cryogenic datasets on the P4(3)2(1)2 form of HEW-Lysozyme.
The structure of the triclinic form of HEW-Lysozyme was solved by direct methods with the program SHELXD [1] - and somewhat later with "Shake & Bake", in both cases using synchrotron data. However, it is still a challenge to try to collect in-house data from the tetragonal form that are good enough for ab initio structure solution.
The best dataset of tetragonal HEW-lysozyme reported in the PDB is 193L and has been measured with synchrotron radiation at room temperature to a resolution of 1.33 A and a completeness of 87.1 % [2]. After optimizing crystallization, cryoconditions and data collection it was possible for us to collect laboratory data at 150 K on a rotating anode / multiwire proportional counter system up to 1.30 A with a completeness of 99.9 %.
In this context it should be emphasized that the quality of data is highly influenced by the redundancy. Although a higher redundancy tends to increase the merging R-value Rsym [3], it also improves the quality and I/sigma ratio for the averaged reflections.
In the last months three datasets from three different crystals have been collected, and a refinement with SHELXL [4] has been performed. At the end of the refinement our model has been compared to all HEW-Lysozyme models deposited to the PDB after the year 1986 with a resolution of 2.1 A or better. Several interesting variations could be observed and in some cases explained.
In the future we would be interested in collecting highly redundand data with various types of diffractometers in order to see how different geometries, monochromator systems, collimators, and other facilities can influence the quality and the resolution of large molecule single crystal x-ray data.