Page 192 - Macromolecular Crystallography

HIGH-THROUGHPUT DATA COLLECTION AT SYNCHROTRONS 181

(a) (b)

Figure 12.4 SGX-CAT automated scoring. (a) Diffraction from a crystal of a protein kinase/ligand complex. Automated crystal scoring
suggested that a dataset should be acquired from this sample. Standard visual inspection would almost certainly reject such a crystal.
(b) Experimental electron density map of the same crystal generated automatically from diffraction data without human intervention. The
presence of a ligand in the active site is readily apparent. The resulting structure was used by the SGX structure-based drug discovery team
to guide further optimization of ligand affinity.

collection. In rare instances of unacceptable sam- unlike a home source, the energy that best serves the
ple degradation, manually optimized data collection diffraction experiment must be selected.
is invoked. If the diffraction experiment does not require a
Automated crystal quality evaluation permits particular wavelength, such as for structure deter-
rapid, objective assessment of each crystal, deter- minations with molecular replacement (Chapter 7),
mining whether or not it is worthy of data collection. convenient choices for the wavelength are 1 Å or
In addition, quantitative comparisons among dupli- 0.9793 Å, the wavelength corresponding to the K
cate crystals improves throughput by focusing data absorption edge of selenium. Such X-ray wave-
collection on only the best crystals. To automate our lengths are readily available at most synchrotron
selection process, a link between replicate crystals beamlines.
must exist within the software control system. At Measurements that rely on anomalous scatter-
SGX, this link is made via the database (Section 12.9). ing, however, are usually most successful when
This combination of software linkage and automatic the X-ray energy matches that of the absorption
scoring has had a significant impact on data col- edge of the anomalous element. Because of concerns
lection throughput. Currently, approximately 70% that the chemical environment of the anomalous
of the samples examined at SGX-CAT are evalu- atoms affects the precise location of the absorp-
ated without manual examination of the diffraction tion edge (Smith and Thompson, 1998), it has been
images, either during screening or data collection. customary to measure the absorption spectrum of
We expect this proportion to increase as further the anomalous element within the protein crys-
software enhancements are made. tal itself. In doing so, the crystal is exposed to
X-rays prior to the diffraction experiment, which can
12.5 Selection of X-ray wavelength damage the crystal. However, many of the effects
of atomic environment and oxidation state on
Akey advantage of synchrotron sources is their abil- the precise location of an absorption edge are
ity to provide X-rays over a wide range of energies well-understood (Koningsberger and Prins, 1988).
or wavelengths. When using synchrotron radiation, Knowledge from the field of X-ray Absorption Near