Abstract
DatView is a new graphical user interface (GUI) for plotting parameters to explore correlations, identify outliers and export subsets of data. It was designed to simplify and expedite analysis of very large unmerged serial femtosecond crystallography (SFX) data sets composed of indexing results from hundreds of thousands of microcrystal diffraction patterns. However, DatView works with any tabulated data, offering its functionality to many applications outside serial crystallography. In DatView's user-friendly GUI, selections are drawn onto plots and synchronized across all other plots, so correlations between multiple parameters in large multi-parameter data sets can be rapidly identified. It also includes an item viewer for displaying images in the current selection alongside the associated metadata. For serial crystallography data processed by indexamajig from CrystFEL [White, Kirian, Martin, Aquila, Nass, Barty & Chapman (2012 ▸). J. Appl. Cryst. 45, 335-341], DatView generates a table of parameters and metadata from stream files and, optionally, the associated HDF5 files. By combining the functionality of several commonly needed tools for SFX in a single GUI that operates on tabulated data, the time needed to load and calculate statistics from large data sets is reduced. This paper describes how DatView facilitates (i) efficient feedback during data collection by examining trends in time, sample position or any parameter, (ii) determination of optimal indexing and integration parameters via the comparison mode, (iii) identification of systematic errors in unmerged SFX data sets, and (iv) sorting and highly flexible data filtering (plot selections, Boolean filters and more), including direct export of subset CrystFEL stream files for further processing.
Highlights
Serial femtosecond crystallography (SFX) at X-ray freeelectron lasers (XFELs) is a recent innovation in macromolecular crystallography where structures are determined from thousands of snapshot diffraction patterns from nano/ microcrystals delivered to a pulsed XFEL beam in a serial fashion
Data visualization is an important guide for this optimization, as well as for optimal data collection
The preprocessing script and configuration file for Pandas HDF5 spectroscopy files at PAL-XFEL are shipped with DatView, and NSLS-II files are linked to from DatView’s GitHub page
Summary
Serial femtosecond crystallography (SFX) at X-ray freeelectron lasers (XFELs) is a recent innovation in macromolecular crystallography where structures are determined from thousands of snapshot diffraction patterns from nano/ microcrystals delivered to a pulsed XFEL beam in a serial fashion. DatView’s unified GUI increases the power of the program to explore correlations between many parameters simultaneously, export selections and compare the effects of different processing approaches It loads any text table where each column is a parameter and each row an item. The external configuration file describes the data type of each parameter so it can be loaded by NumPy, and is necessary for any data containing non-numeric fields It allows the configuration of many other parameters such as color maps, default histograms and displays in the item viewer. The preprocessing script and configuration file for Pandas HDF5 spectroscopy files at PAL-XFEL are shipped with DatView, and NSLS-II files are linked to from DatView’s GitHub page
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
