JUNGFRAU detector for brighter x-ray sources: Solutions for IT and data science challenges in macromolecular crystallography.

Filip Leonarski,Bernd Schmitt,Carlos Lopez-Cuenca,Oliver Bunk,Heinrich Billich,Martin Brückner,Sophie Redford,Leonardo Sala,Aldo Mozzanica,Andrej Babic,Meitian Wang

doi:10.1063/1.5143480

Filip Leonarski, Bernd Schmitt + Show 9 more

Open Access

https://doi.org/10.1063/1.5143480

Copy DOI

Abstract

In this paper, we present a data workflow developed to operate the adJUstiNg Gain detector FoR the Aramis User station (JUNGFRAU) adaptive gain charge integrating pixel-array detectors at macromolecular crystallography beamlines. We summarize current achievements for operating at 9 GB/s data-rate a JUNGFRAU with 4 Mpixel at 1.1 kHz frame-rate and preparations to operate at 46 GB/s data-rate a JUNGFRAU with 10 Mpixel at 2.2 kHz in the future. In this context, we highlight the challenges for computer architecture and how these challenges can be addressed with innovative hardware including IBM POWER9 servers and field-programmable gate arrays. We discuss also data science challenges, showing the effect of rounding and lossy compression schemes on the MX JUNGFRAU detector images.

Highlights

Macromolecular crystallography (MX) is the dominant method for high-resolution structure determination of biomolecules
In this paper, we present a data workflow developed to operate the adJUstiNg Gain detector FoR the Aramis User station (JUNGFRAU) adaptive gain charge integrating pixel-array detectors at macromolecular crystallography beamlines
We summarize current achievements for operating at 9 GB/s data-rate a JUNGFRAU with 4 Mpixel at 1.1 kHz frame-rate and preparations to operate at 46 GB/s data-rate a JUNGFRAU with 10 Mpixel at 2.2 kHz in the future

Summary

INTRODUCTION

Macromolecular crystallography (MX) is the dominant method for high-resolution structure determination of biomolecules. The anomalous peak height, the precision of determining the last shell, and refinement statistics are all affected by lossy compression, but such a difference could be acceptable for some applications Such compression could be used in parallel with rounding schemes—an SZ compressed dataset could be taken home on a portable hard disk even for the most data 1⁄4 intensive serial crystallography experiments, for example for the ability to visualize images at user’s home institution, while the higher precision data would remain available for processing in the high performance computing center at the synchrotron/XFEL facility. It will be only possible to operate above a certain frame rate if online data reduction is implemented by either a rounding scheme, lossy compression, or a veto mechanism for empty images in serial crystallography Such choices should be made so that the imprecision and information loss in the data have a negligible impact on the outcome of the experiment. The gain in the sustainable data rate will be beneficial for high-throughput applications and could enable scientists to better explore what the next-generation light sources can offer

METHODS

Findings

CODE AVAILABILITY