<i>KAMO</i>: towards automated data processing for microcrystals

TLDR

Radiation damage limits high‑resolution data from single protein microcrystals, so collecting small wedges from many crystals and merging them is promising but difficult to process. KAMO was created as an open‑source pipeline to automate the entire processing of multiple small‑wedge datasets. It processes individual datasets, groups those with matching unit cells, selects the space group, resolves indexing ambiguities, clusters, merges with outlier rejection, and generates a report. Tests on synthetic and real data from hundreds of crystals show that KAMO can automatically merge structure‑factor amplitudes, greatly aiding structure determination of challenging microcrystal targets.

Abstract

In protein microcrystallography, radiation damage often hampers complete and high-resolution data collection from a single crystal, even under cryogenic conditions. One promising solution is to collect small wedges of data (5-10°) separately from multiple crystals. The data from these crystals can then be merged into a complete reflection-intensity set. However, data processing of multiple small-wedge data sets is challenging. Here, a new open-source data-processing pipeline, KAMO, which utilizes existing programs, including the XDS and CCP4 packages, has been developed to automate whole data-processing tasks in the case of multiple small-wedge data sets. Firstly, KAMO processes individual data sets and collates those indexed with equivalent unit-cell parameters. The space group is then chosen and any indexing ambiguity is resolved. Finally, clustering is performed, followed by merging with outlier rejections, and a report is subsequently created. Using synthetic and several real-world data sets collected from hundreds of crystals, it was demonstrated that merged structure-factor amplitudes can be obtained in a largely automated manner using KAMO, which greatly facilitated the structure analyses of challenging targets that only produced microcrystals.