Rugnux: from diffraction images to structure at synchrotrons and XFELs in one program

Rugnux converts rotation or serial still diffraction images into merged data and an initial structure, delivering synchrotron results within seconds and processing 207 public datasets with a 15 s median per dataset.

• Why it matters: Growing macromolecular crystallography data rates strain multi-program cluster workflows, creating delays between data collection and usable structures.
• What they did: The single GPU-accelerated program processes diffraction images into merged datasets and an initial model, tested at synchrotrons and on serial XFEL data.
• The result: At synchrotrons, outputs are produced seconds after the final image, and serial XFEL processing can run faster than collection, enabling real-time structure determination.

AI summary of the abstract · may contain errors · not medical advice · read the paper.