Processing serial synchrotron crystallography diffraction data with DIALS

Abstract

This chapter describes additions to the DIALS software package for processing serial still-shot crystallographic data, and the implementation of a pipeline, xia2.ssx, for processing and merging serial crystallography data using DIALS programs. To integrate partial still-shot diffraction data, a 3D gaussian profile model was developed that can describe anisotropic spot shapes. This model is optimised by maximum likelihood methods using the pixel-intensity distributions of strong diffraction spots, enabling simultaneous refinement of the profile model and Ewald-sphere offsets. We demonstrate the processing of an example SSX dataset where the improved partiality estimates lead to better model statistics compared with post-refined isotropic models. We also demonstrate some of the workflows available for merging SSX data, including processing time/dose resolved data series, where data can be separated at the point of merging after scaling and discuss the program outputs used to investigate the data throughout the pipeline.

Keywords: DIALS; Data integration; Data processing; Data reduction; Serial crystallography; Xia2.ssx.