October 1999 issue
Data collection and processing
Proceedings of the CCP4 study weekend
Key features of the applications of the macromolecular crystallography beamlines at the ESRF are described.
Properties such as the required size, divergence, wavelength spread and intensity of the X-ray beam, together with the size and resolution of the detector, are derived from the properties of protein crystals and their diffraction patterns.
Charge-coupled device detectors are compared with other commercially available detectors and their use with home X-ray sources is discussed.
The performance of the novel MicroSource X-ray generator is analysed and predictions are made for alternative designs of focusing optics. Relative performance of this microfocus tube and an optimized mirror system become increasingly advantageous with the study of ever-smaller crystals.
A detailed comparison of usable flux, spectral purity, divergence, beam profile and data quality for home-laboratory X-ray optics systems (total-reflection mirrors and multilayer monochromators) is presented.
A one-dimensional FFT indexing routine developed previously is discussed with particular reference to its implementation in MOSFLM. The method has been shown to be robust and reliable even for unfavourable test images.
The expectations and consequences of the processing of diffraction images with thick and thin rotation-angle increments are discussed. The d*TREK suite for processing images is briefly introduced.
Deconvolution of overlapped spots, careful integration of weak spots and the use of statistical analysis tools are shown to improve the data quality of monochromatic data sets. The described techniques have been implemented in the integration software PrOW.
Cryo-electron microscopy cryo-EM) and crystallography are complimentary tools now available to the structural biologist. A brief discussion of the problems faced, and the advantages of cryo-Em and three-dimensional image reconstruction, along with three case studies, will be discussed.
The detection and analysis of diffraction data from twinned macromolecular crystals is discussed, as is the recovery of useful data from these crystals.
Although it is better, if possible, to collect highly redundant diffraction data, the most damaging outliers can be detected using structure-factor probability distributions.
Experiments with macromolecular microcrystals at the EMBL/ESRF have shown that this technique offers several advantages. A custom-designed microdiffractometer is being built to ease microcrystal handling.