Jean Jakoncic

The paper evaluates lossy compression methods for high data rate macromolecular crystallography to minimize information loss while maintaining structural analysis quality.

The paper introduces a multi-stage clustering strategy for serial crystallography data to improve dataset merging accuracy by first clustering based on unit cell parameters and then on reflection intensities.

The paper introduces a multi-factor clustering method to classify diffraction data from dynamic proteins into polymorphs, using unit cell constants and intensities to cluster mixed-state data from chymotrypsinogen crystals.