HKL-2000 v718 distribution was released for Linux, Linux 32 and Mac
Version 717 is the first release of HKL with an automatic mode, further simplifying the process of determining a structure. Both HKL-2000 and HKL-3000 can process and scale your data, and HKL-3000 can determine and build your structureautomatically. The "Auto" tab allows automatic indexing, integration, and scaling of data using the hkl_auto program. In benchmark tests on a batch of several thousand data sets, the automatic mode had the success rate of ~83% for indexing and integration and ~76% for space group determination (success was determined by comparison with processing reported in PDB). The automatic mode uses the CCP4 program "pointless" for final space group determination (CCP4 setup needs to be sourced before running HKL). In HKL-3000, the "Auto" feature of the Structure tab will attempt to phase your data and build your model. Please note that some formats or constructs supported in the manual mode are not yet available in the automatic mode.
Other new features of release 717 include the new options to invert beam position read from the frame header. Beam position options have been added to "Site Configuration" window in the "Advanced" tab. You may now 1) always read the beam position from the header or 2) invert it. Also, the 3D Window (the number of frames that will be processed together for indexing and integration) is now calculated based on the frame width and memory limits of the denzo_3d program.
The latest version 716.1 can handle data collected by the new Eiger 16M detector at the Shanghai Synchrotron Research Facility, beamline 17U1.
New features include:
- "Site included" option allows one to place the def.site file in the data (frames) or processing directories - instead of in a directory in the hklint sites directory. This option is available in the initial site selection window,
- scalepack with auto corrections generates 6 additional scaled (.sca) files that are tuned for various tasks like: displaying radiation damage and anomalous maps, structure refinement, generating statistics,
- Xdisp Qt is now the default frame display program. Old Xdisp is available in the 'Options/XDisp/XDisp X' main menu,
- "Check Space Group" is integrated with the "pointless" program from CCP4 and allows to apply the selected space group for the next scaling. It also auto generates required re-indexing matrix,
- new 'Merge' tab allows combining 2 or more .sca files together.
The changes from HKL-2000 v714 to v715 are:
- handling of NeXus (.h5) data was improved. Read times are around 2x faster - support for variable 32-bit/16-bit data was added to NeXus format (.h5). This resolved issues in earlier versions (Xdisp Qt displayed black bars and higher percentage of reflections were rejected during scaling)
- Linux 32-bit distribution now works with NeXus lz4 bitshuffle format
- option to read beam position from frame header was moved to the 'Advanced' tab of the 'Site Configuration' window
- reading the beam position from frame header was enabled for NeXus (.h5) format
- Rigaku PAD4M, Rigaku Oxford Atlas, Rayonix mx170hs and Pilatus 1M detector formats were added
- indexing and integration handling procedures were improved
- 'Short Version' option was added to scaling report. It does not include tables for integration statistics
- reindexing window was updated
- option to keep original lattice symmetry was added to 'Check Space Group' window
- input to 'Check Space Group' was updated to handle primitive orthorhombic groups to keep "unique" axis along c.
Diffraction images from Eiger detectors are written as .h5 files (NeXus format), in which multiple images may be compressed into one file. HKL-2000 release 714 can process .h5 files directly, thereby sidestepping the need to extract images manually and convert them into individual .cbf files. From the user's perspective, processing is handled just like any other file format (.cbf, .img, etc).
The newest releases of HKL-2000 and HKL-3000 include a new, optional version of the program that displays diffraction images - XDisp Qt.
Major advantages of XDisp Qt:;
- Better navigation and scalable window
- Zoom in the main window using the mouse wheel
- Option to merge frames for inspection/peak search
- Easier selection of the beam position
- Easy, intuitive tool for editing the blind/shadow regions
- Increased control of display contrast using a histogram / slider combination.
To use this new version, select "Options -> XDisp -> XDisp Qt
HKL-2000 v714 has been released.
It includes new optional version of frame display program - XDisp Qt. By default the old XDisp will be used.
To use the new one, in the main menu, select 'Options --> Xdisp --> XDisp Qt'.
The selection can be saved for the next sessions using 'Options --> Save options'.
Other changes include:
- Initial support for h5 files for Pilatus and Eiger detectors was added (not supported in 32-bit Linux)
- Support for the Rayonix MX300HE format was added
- Support for the Stoe pad200k format was added
- The Pilatus “read header” procedure was updated
- The “split data sets” window was updated
- A button that shows scaling statistics for the selected .sca file was added to the 'Project' tab
- The program is now distributed as a compressed directory that includes all binaries, related data, and library files
Distribution v712 for the HKL-2000 suite was released on March 24, 2016. The new version includes support for the Stoe pad 300k detector and the horizontal Euler goniostat. A new 'Select All' option was added to the dataset selection dialog. Other enhancements include improvements to the handling of the Eiger 9m detector and the cbf headers, as well as an improved dataset template generation procedure. Additionally, the scalepack input file will now be saved using the current .sca file name template.
Using Mac OSX 10.10 and 10.11, the initial rendering of each window is slow, but once the interface is loaded it works normally. Additionally, if the ‘Peak Search’ button on the Index page is used before the image has been displayed, the Xdisp window will close. The workaround for this is to first display the image using the use ‘Display’ button prior to performing a peak search. See the Hardware & Operating Systems page for more details.