Detecting ice artefacts in processed macromolecular diffraction data with machine learning

Contamination with diffraction from ice crystals can negatively affect, or even impede, macromolecular structure determination, and therefore detecting the resulting artefacts in diffraction data is crucial. However, once the data have been processed it can be very difficult to automatically recognize this problem. To address this, a set of convolutional neural networks named Helcaraxe has been developed which can detect ice-diffraction artefacts in processed diffraction data from macromolecular crystals. The networks outperform previous algorithms and will be available as part of the AUSPEX web server and the CCP4-distributed software.

The crystal structure of AbsH3: A putative flavin adenine dinucleotide-dependent reductase in the abyssomicin biosynthesis pathway

Natural products and natural product-derived compounds have been widely used for pharmaceuticals for many years, and the search for new natural products that may have interesting activity is ongoing. Abyssomicins are natural product molecules that have antibiotic activity via inhibition of the folate synthesis pathway in microbiota. These compounds also appear to undergo a required [4 + 2] cycloaddition in their biosynthetic pathway. Here we report the structure of an flavin adenine dinucleotide-dependent reductase, AbsH3, from the biosynthetic gene cluster of novel abyssomicins found in Streptomyces sp. LC-6-2.

Background modelling of diffraction data in the presence of ice rings

An algorithm for modelling the background for each Bragg reflection in a series of X-ray diffraction images containing Debye-Scherrer diffraction from ice in the sample is presented. The method involves the use of a global background model which is generated from the complete X-ray diffraction data set. Fitting of this model to the background pixels is then performed for each reflection independently. The algorithm uses a static background model that does not vary over the course of the scan. The greatest improvement can be expected for data where ice rings are present throughout the data set and the local background shape at the size of a spot on the detector does not exhibit large time-dependent variation. However, the algorithm has been applied to data sets whose background showed large pixel variations (variance/mean > 2) and has been shown to improve the results of processing for these data sets. It is shown that the use of a simple flat-background model as in traditional integration programs causes systematic bias in the background determination at ice-ring resolutions, resulting in an overestimation of reflection intensities at the peaks of the ice rings and an underestimation of reflection intensities either side of the ice ring. The new global background-model algorithm presented here corrects for this bias, resulting in a noticeable improvement in R factors following refinement.

AUSPEX: a graphical tool for X-ray diffraction data analysis

In this paper, AUSPEX, a new software tool for experimental X-ray data analysis, is presented. Exploring the behaviour of diffraction intensities and the associated estimated uncertainties facilitates the discovery of underlying problems and can help users to improve their data acquisition and processing in order to obtain better structural models. The program enables users to inspect the distribution of observed intensities (or amplitudes) against resolution as well as the associated estimated uncertainties (sigmas). It is demonstrated how AUSPEX can be used to visually and automatically detect ice-ring artefacts in integrated X-ray diffraction data. Such artefacts can hamper structure determination, but may be difficult to identify from the raw diffraction images produced by modern pixel detectors. The analysis suggests that a significant portion of the data sets deposited in the PDB contain ice-ring artefacts. Furthermore, it is demonstrated how other problems in experimental X-ray data caused, for example, by scaling and data-conversion procedures can be detected by AUSPEX.

De-icing: recovery of diffraction intensities in the presence of ice rings

Macromolecular structures are routinely determined at cryotemperatures using samples flash-cooled in the presence of cryoprotectants. However, sometimes the best diffraction is obtained under conditions where ice formation is not completely ablated, with the result that characteristic ice rings are superimposed on the macromolecular diffraction. In data processing, the reflections that are most affected by the ice rings are usually excluded. Here, an alternative approach of subtracting the ice diffraction is tested. High completeness can be retained with little adverse effect upon the quality of the integrated data. This offers an alternate strategy when high levels of cryoprotectant lead to loss of crystal quality.