Otwinowski Z. and Minor W., Processing of X-ray Diffraction Data Collected in Os

- Protein crystallography

Processing of X-ray Diffraction Data Collected in Oscillation Mode, Methods in Enzymology, Volume 276: Macromolecular Crystallography, part A, p.307-326, 1997,C.W. Carter, Jr. & R. M. Sweet, Eds.,

Otwinowski Z. and Minor W.

Department of Molecular Physiology and Biological Physics, University of Virginia, Charlottesville, VA 22903, USA.

Introduction

X-ray data can be collected with 0-, 1-, and 2-dimensional detectors, 0-D (single counter) being the simplest and 2-D the most efficient in terms of rneasuring diffracted X-rays in all directions. Two-dimensional detectors have been used since 1912 for X-ray diffraction studies. Initially the 2-D detector was made of X-ray sensitive photographic film; now electronic detectors and phospholuminescent films (best known by the trade name IP or Imaging Plate) dominate. To analyze single-crystal diffraction data collected with these detectors, several computer programs have been developed. The 2-D detectors and related software are now used predominantly to measure and integrate diffraction from single crystals of biological macromolecules. However, the usefulness of these systems in small-molecule, high-resolution crystallography is just being recognized and much of the rest of this discussion is applicable to that field also...

Summary

Macromolecular crystallography is an iterative process. Rarely do the first crystals provide all the necessary data to solve the biological problem being studied. Each step benefits from experience learned in previous steps. To monitor the progress, the HKL package provides two tools:

Statistics, both weighted (chi²) and unweighted (R_merge), are provided. The Bayesian reasoning and multicomponent error model facilitates the obtaining of proper error estimates (Schwarzenbach:40) and
Visualization of the process plays a double role: it helps the operator to confirm that the process of data reduction, including the resulting statistics, is correct, and it allows one to evaluate problems for which there are no good statistical criteria. Visualization also provides confidence that the point of diminishing returns in data collection and reduction has been reached. At that point the effort should be directed to solving the structure.

The methods presented here have been applied to solve a large variety of problems, from inorganic molecules with 5 Å unit cell to rotavirus of 700 Å diameter crystallized in 700 x 1000 x 1400 Å cell (Temple:41). Overall quality of the method has been tested by many researchers by successful application of the programs to MAD structure determinations.