Processing X-ray free-electron laser (XFEL) diffraction images poses challenges, as an XFEL pulse is powerful enough to destroy or damage the diffracting volume and thereby yields only one diffraction image per volume. Moreover, the crystal is stationary during the femtosecond pulse, so reflections are generally only partially recorded. Therefore, each XFEL diffraction image must be scaled individually and, ideally, corrected for partiality prior to merging. An additional complication may arise owing to indexing ambiguities when the symmetry of the Bravais lattice is higher than that of the space group, or when the unit-cell dimensions are similar to each other. Here, an automated method is presented that diagnoses these indexing ambiguities based on the Brehm–Diederichs algorithm [Brehm & Diederichs (2014), Acta Cryst. D70, 101–109] and produces a consistent indexing choice for the large majority of diffraction images. This method was applied to an XFEL diffraction data set measured from crystals of the neuronal SNARE–complexin-1–synaptotagmin-1 complex. After correcting the indexing ambiguities, substantial improvements were observed in the merging statistics and the atomic model refinement R values. This method should be a useful addition to the arsenal of tools for the processing of XFEL diffraction data sets.
Femtosecond free-electron laser x-ray diffraction data sets for algorithm development