A novel procedure for the characterization and the arithmetic classification of monoatomic multilattices in any dimension is proposed. The algorithm may be coded to provide an automatic classification procedure.

A predicted model-aided one-step classification–multireconstruction algorithm for X-ray free-electron laser single-particle imaging is proposed. The algorithm is capable of processing mixed diffraction patterns from multiple molecules, classifying diffraction patterns by different molecules, determining their orientations and reconstructing multiple 3D diffraction intensities, in one step.

Many models have been developed for analyzing SAXS data; however choosing the optimal model is difficult and time-consuming, especially for non-expert users. This paper proposes an algorithm, based on machine learning, representation learning and SAXS-specific preprocessing methods, which instantly selects the nanoparticle model best suited to describe SAXS data.

STOPGAP, an open-source package for subtomogram averaging that is designed to provide users with fine control over each of the steps in the image-processing workflow, is described.

A curve-tracking computer algorithm to classify crystallization images automatically is described.

An efficient and robust algorithm for the comparison of clusters is presented. Several illustrative example applications are given, including the screening of sets of clusters generated during global optimizations and Monte Carlo/molecular dymanics simulations, the identification of specific structure fragments inside large clusters, and the study of structure-substructure relations of periodic crystals.

A new artificial-intelligence-based platform, CrystalMELA, that can implement machine-learning models has been developed. Powder X-ray diffraction patterns of organic, inorganic and metal–organic compounds and minerals were used to train and test the learning models, and CrystalMELA has been employed for crystal system classification.

This paper proposes a pipeline to categorize serial crystallography data, consisting of a real-time feature extraction algorithm, an image descriptor and a machine learning classifier. This approach demonstrates superior performance compared with other feature extractors and classifiers.

Irreducible representations can now be used to derive the isotropy subgroups at irrational wavevectors, making it relatively easy to explore the (3 + d)-dimensional superspace-group symmetries that arise from incommensurate modulations of a parent crystal structure. A general algorithm capable of arbitrary superpositions of multiple incommensurate and commensurate order parameters is presented.

A novel strategy is presented for sample jet alignment using machine vision for liquid-jet-based sample delivery systems. Feedback using height-resolution images from an optical microscope positioned perpendicular to the path of the X-ray beam enables tracking of the relative alignment of the liquid jet and X-ray beam.