A step towards making serial femtosecond X-ray crystallography data processing user-friendly.

What is it about?

Serial femtosecond crystallography (SFX) is a technique for determining the structure of biological molecules at atomic resolution. However, this technique requires highly intense, ultra-short X-ray pulses, experiment-specific methods of introducing the biological samples into the X-ray beam, and novel detectors for capturing the data. This results in often complicated analysis for extracting useful images, and the Pixel Anomaly Detection Tool (PADT) provides a user-friendly interface that makes this analysis amendable to the lay programmer by making use of established machine learning algorithms.

Featured Image

Why is it important?

An SFX experiment often requires a large team of scientists with various expertise, while generating large amounts of image data (often multiple terrabytes). From sample preparation, sample injection, X-ray instrumentation to data analysis; everything influences the collected data and different scientists are interested in studying different effects. PADT provides an interactive interface to easily train machine learning models for filtering out the images containing the requested information, without having to write any code.

Perspectives

“

Our goal was to create a tool that would make advanced data sorting and analysis of SFX data accessible to researchers without extensive programming expertise. We believe that PADT achieves this goal, and further enhances the quality of data analysis. We hope that the friendliness of encapsulating this in a graphical user interface will have a positive effect on the amount of time and effort spent on cleaning up XFEL data and/or data reduction.

”