Loos-4.1.0

This release includes a number of fixes related to issues listed on github. Added support for PDBx, mmCIF, and MDTraj-HDF5 formats.

Loos-4.0.4

This release includes a number of fixes that address recent issues listed on github. Molecular
order parameter function was also added to AtomicGroup.

Loos-4.0.3

This is another maintenance release that addresses using the correct OpenBLAS/lapack libraries, corrects the version number reported by LOOS, and includes a formatting change for the ramachandran tool.

Loos-4.0.2

This is a maintenance release. It includes a minor change to how loos is built for OS X.

Loos 4.0.1

This is a maintenance release that addresses an issue with using AtomicGroups as iterators in Python.

Loos 4.0

We are pleased to announce the release of LOOS-4.0. If you have been working from the main branch, there should be minimal change in how you use LOOS. However, behind the scenes, there have been many changes. In particular, we are now using cmake rather than scons for building LOOS. In addition, PyLOOS is installed via pip into your Python (or conda) environment. There are also several new tools.

Within the next week or two, we expect to release a version installable via conda-forge, which will make installation much simpler.

LOOS 3.3

This release has a new and improved facility for reweighting (from Louis Smith), methods for calculating logistic and hard-cutoff lateral densities in AtomicGroup, several new tools for working with clusters in the Clustering package, and dipole computation in membrane_map.

There are also a number of bugfixes:

• Incorrect bonds removed from add_molecule.py
• Fixed units for periodic box read by gmxdump2pdb.pl
• Update to handle new constraints format in gromacs 2020.3+ in gmxdump2pdb.pl
• Fixed charge units when reading Amber prmtop files
• OMG correctly handles 4- and 5-site water models
• In subsetter and merge-traj, the --regex and --scanf flags now imply --sort

There were also a few housekeeping updates. We migrated continuous integration from Travis-CI to GitHub Actions, Louis Smith contributed a much cleaner .gitignore file, and we fixed a potential build-bug where the C++ file generated by swig had different names depending on which compiler/swig version was used. We also added scikit-learn to the list of conda packages installed by conda_build.sh, and fixed a silly bug in its detection of the available compilers.

LOOS 3.2

This is a major new release of LOOS, with several major changes:

1. The install has been further stream-lined, especially under Conda. If you build and install in a conda environment, the binaries get installed in the conda bin directory, and the python libraries in the Conda's python site-packages. Under these circumstances, it is no longer necessary to source setup.sh to get paths set correctly. As part of this work, the location of the Voronoi and OptimalMembraneGenerator packages has changed -- any code that imports these packages will need to be updated.

2. Several new tools:
   a. Better radius of gyration tool rad-gyr (Louis Smith)
   b. New electrostatic potential mesh tool esp_mesh for charge fitting (Chapin Cavender)
   c. Track multiple dihedral angles through a trajectory with dihedrals (Louis Smith)
   d. Categorize RNA structure using ran_suites (Chapin Cavender)
   e. Added experimental tSNE option to cluster-structures.py (requires scikit-learn)

3. Multiple bug fixes and refinements:
   a. Bug fixed in svd tool, where splitting the v-matrix wasn't done correctly when invoked with multiple trajectories.
   b. aligner tool warns if you choose the --append option with iterative alignment
   c. membrane_map can now run without aligning, useful for tracking membrane domains

LOOS 3.1

This is a major new release. Most importantly, the build system has been totally reworked, so compiling LOOS should be much easier. The preferred approach is to use conda, which (as far as we can tell) works on all linux and mac systems.

v3.0.0

This is a major release of LOOS. The biggest changes are:

• PyLOOS and our build system have switched from python 2.7 to python 3.x
• a new reweighting facility
• lots of small bug fixes