AN ANALYSIS OF DATA COLLECTION STRATEGIES AND DATA REDUCTION SOFTWARE FOR IMAGE PLATE DATA. Bing Hu¹, John Rose² and Bi-Cheng. Wang², ¹Dept. of Crystallography, University of Pittsburgh, Pittsburgh, PA 15260, U.S.A., ²Dept. of Biochemistry & Molecular Biology, University Georgia, Athens, GA 30602, U.S.A.

In designing a strategy for macromolecular data collection, one of the most often asked questions is "What scan range should I use?". The answer may vary depending on the facilities used for data collection and the programs used for data processing. Intituitively, narrow oscillation data slices (0.1-0.25deg.) should give a better signal-to-noise ratio than large oscillation data slices (say, 0.5-2.0deg.). However most image plate data is collected using a large oscillation data slice due in part to the slow readout time of the commerical detectors, the deacy of diffraction quality with time and the initial lack of data reduction software for narrow oscillation data slices. In addition, there appears to be no reported systematic study on the relation of scan range to data quality for image plate data which prompted us to do a systematic study on this subject. Since data quality may also be affected by the techniques (e.g. 2D versus 3D profile fitting) used in the data reduction, we have included data quality versus data reduction software used as part of our study.

The data used in this study was a 2.6Å set of 0.25deg. oscillation data collected at room temperature using a Mar Research 30 cm image plate scanner on crystals of the Neurophysin-hydrin I complex (space group P4₁2₁2, a=b=68.7Å and c=113.64Å). The 0.25( images were combined to form the 0.5(, 0.75deg., 1.00deg., 1.25( and 1.5deg. used in the analysis. Each data set was then processed using X-GEN, XDS, MOSFILM and DENZO. The results of our analysis will be presented.

Work supported by NIH grant GM-41936.