gorder v1.4.1

gorder v1.4.1

Changes in version 1.4.1

Bug fixes

• Fixed a bug where the collect/export keyword would not be parsed correctly if its value was false.

gorder v1.4.0

gorder v1.4.0

Changes in version 1.4.0

Inverting geometric selections

• Geometry selections can be now inverted using the invert keyword. See the manual for more details.

Other changes

• A companion tool GUIorder providing graphical user interface for gorder has been released. See the manual for more details.
• gorder along with its documentation has been transferred to Vacha Lab organization.

gorder v1.3.0

gorder v1.3.0

Changes in version 1.3.0

Faster leaflet assignment for vesicles

• Implemented a new leaflet classification method, SphericalClustering, exclusively for vesicles. This method classifies lipids using a 2-component Gaussian Mixture Model based on the distances between lipid headgroups and the vesicle center of geometry. This method is much faster than Clustering, but is only usable for systems with a single unilamellar vesicle. See the manual for more details.

Analyzing very long trajectories

• Trajectories longer than 2 147 483 647 simulation steps can be now analyzed.

gorder v1.2.0

gorder v1.2.0

Changes in version 1.2.0

Exporting internal data

• You can now collect and export information about which leaflets individual lipids were assigned to during the analysis, as well as dynamically calculated membrane normals, to external files. This is only performed when requested in the analysis setup. The data can also be accessed via the Rust and Python APIs. Read more in the corresponding sections of the manual: Exporting lipid assignment data and Exporting membrane normals.

Flipping the assignment

• When assigning lipids to leaflets, you can use the flip keyword, which reverses the assignment—upper leaflet becomes lower leaflet and vice versa. This may seem useless, but it can help with leaflet assignment when using the clustering method. See the manual for more details.

Python API

• The Python gorder library now exposes type information and supports autocomplete.
• Documentation of the Python API has been improved and should be much more complete.

Bug fixes

• Manually assigning membrane normals previously returned an error unless the analysis step was set to 1. This has been fixed, and manual membrane normals can now be used with any analysis step.

Other changes

• Changed the wording of some error messages and adjusted the formatting of certain output written to stdout.

gorder v1.1.0

gorder v1.1.0

New definition of the @membrane macro.

Changes in version 1.1.0

• The @membrane macro now matches a wider variety of lipid types. See the GSL guide for details on how the macro currently expands.
  • This change makes selecting lipid molecules that form your membrane simpler and more reliable, especially when working with less common lipid types.
  • The macro still cannot guarantee identification of all lipid molecules, so exercise caution when using it.
  • Since the @membrane macro now matches more lipid types, some queries using this macro may return different atom selections. Therefore, gorder v1.1.0 is not guaranteed to return the same results as gorder v1.0.0 or previous versions.
  • Example of behavior change and required modification: In the gorder article (Figure 3), we used the query (@membrane and name r'C3.+|C2.+') or (resname PVCL2 and name r'C[ABCD].+') to select all lipid carbons in a complex system. In previous versions of gorder, PVCL2 was not identified as a membrane lipid when using the @membrane macro. This behavior has now changed, and the PVCL2 molecule is matched by the @membrane macro. In gorder v1.1.0, the above query should therefore be modified to (@membrane and not resname PVCL2 and name r'C3.+|C2.+') or (resname PVCL and name r'C[ABCD].+') (this is one of the rare cases where the update complicates things, as we don't want the C3 and C2 atoms of PVCL2 to be selected).

gorder v1.0.0

First stable version of gorder.

Changes in version 1.0.0

• Support for GROMACS 2025 TPR files.
• More precise calculation of geometric center of groups using the Refined Bai-Breen algorithm. Note that as a consequence of this, gorder 1.0.0 may return slightly different results for some systems than previous versions.
• Providing the current working directory as the directory for saving ordermaps will now result in an error.
• Changes to the Rust API: Made all user-provided parameters of all methods for leaflet assignment public.

gorder v0.7.1

Version 0.7.1 of gorder.

Changes in version 0.7.1

• Fixed undefined behavior in the groan_rs library's C source code that caused TRR files to be read incorrectly when compiled with certain C compilers (e.g., clang).

gorder v0.7.0

Version 0.7 of gorder: Leaflet classification for curved membranes

Changes in version 0.7

• Leaflet classification for curved membranes: gorder can now use spectral clustering to classify lipids into leaflets for any membrane geometry, including buckled membranes and vesicles. See the manual for more information.
• !!BREAKING CHANGE!! Removed support for some trajectory formats: Removed support for PDB, Amber NetCDF, DCD, and LAMMPSTRJ trajectories. (PDB structure files are still supported.) The parsing of these trajectories relied on the chemfiles library, whose maintenance standards and capabilities just do not meet the quality requirements for gorder. If you need to use these trajectory formats, you can still use gorder v0.6.

gorder v0.6.0

Version 0.6 of gorder: United-atom order parameters, Python API, and more.

Changes in version 0.6

• United-atom order parameters: gorder is now able to calculate order parameters in united-atom systems. See the manual for more information.
• Python API: gorder is now available as a Python package. See the manual for more information.
• Trajectory concatenation: You can now provide multiple trajectory files which will be all joined into one trajectory and analyzed. In case there are duplicate frames at trajectory boundaries, gorder will analyze only one of the duplicate frames. This feature is currently only supported for XTC and TRR files. See the manual for more information.
• Manual membrane normals: Membrane normals can be now assigned manually for each lipid molecule in each trajectory frame. See the manual for more information.
• Optimizations: Local leaflet classification method and dynamic local membrane normal calculation are now much faster, especially for very large systems, through using cell lists. Warning: The changes may cause small differences in the calculated order parameters compared to version 0.5, especially for very short trajectories. These differences should be on the order of 0.0001 arb. u. or lower.

gorder v0.5.0

Version 0.5 of gorder: More features, more colors

Changes in version 0.5

• Dynamic membrane normal calculation: Membrane normals can be now calculated dynamically from actual membrane shape which allows the calculation of order parameters for vesicles and similar systems (see the manual).
• Ignoring PBC: You can now choose to ignore periodic boundary conditions. This allows analyzing simulations with non-orthogonal simulation boxes with some small additional friction (making molecules whole). See the manual for more information.
• Reworked manual lipid assignment: It is now possible to classify lipids into membrane leaflets using NDX files, enabling integration of gorder with FATSLiM. BREAKING CHANGE The keywords to request manual leaflet assignment using a leaflet assignment file have been changed, see the manual for more information.
• Ordermaps visualization: A python script is now generated inside any created ordermaps directory which can be used to easily plot the ordermaps. Changed the default range of the colorbar in ordermaps to more reasonable values.
• More trajectory formats: Added experimental support for more trajectory formats, namely TRR, GRO, PDB, Amber NetCDF, DCD, and LAMMPSTRJ. Always prefer using XTC trajectory as gorder is optimized to read it very fast. There are also some limitations connected with using different trajectory formats, see the manual.
• Bug fixes and other changes:
  • gorder now returns an error if the center of geometry calculation for leaflet classification is nonsensical (i.e., nan).
  • If molecule classification runs longer than expected, progress is logged. By default, progress output begins after 500 ms, but you can adjust this delay using the environment variable GORDER_MOLECULE_CLASSIFICATION_TIME_LIMIT.
  • Added more color to information written to standard output during analysis. Changed logging for output file writing. Changed logging for molecule types.
  • Reading an ndx file with invalid or duplicated group names no longer results in a hard error but instead raises a warning.