Change Log

0.4.2

Added max_nmoves as a configurable field in SomdConfig and made nmoves/ncycles computed properties derived from runtime, timestep, max_nmoves and energy_frequency.

This prevents memory overflow from single-cycle long runtimes while keeping runtime the single source of truth.

0.4.1

Fixed the statistical inefficiency timestep units from femtoseconds to nanoseconds.

0.4.0

Note: This is a breaking change! Old pickle files are not compatible with this version. If you need to load old pickle files, please revert to version 0.3.x.

Decoupled SOMD engine from a3fe calculation and removed run_somd.sh and template_config.cfg, replacing them with a3fe.configuration.slurm_config.SlurmConfig and a3fe.configuration.engine_config.SomdConfig.
SystemPreparationConfigs saved as readable yaml file, rather than as a pickle.
Separated system preparation configuration from system preparation and configured SOMD with a3fe.configuration.system_prep_config.SomdSystemPreparationConfig.
Streamlined the installation process with a Makefile.

Migration Guide: In previous versions, users needed to configure run_somd.sh and template_config.cfg files to run calculations. However, with version 0.4.0, these files are no longer needed. Instead, configuration objects (SlurmConfig, SomdConfig, and SomdSystemPreparationConfig) are passed directly to the Calculation object.

For examples and detailed usage of these configuration objects, see the Customising Calculations section in the documentation.

0.3.6

Pinned alchemlyb to version 1.0.1 to avoid compatibility issues with pymbar<4.

0.3.5

Changed the defaults for runtime_constants and thermodynamic speed (“delta_er”) to be consistent with the optimised values in the publication (https://pubs.acs.org/doi/full/10.1021/acs.jctc.4c00806).

0.3.4

Added loguru to make dependency explicit in the environment.yaml and conda-envs/test_env.yaml files.

0.3.3

Fixed bug in water analysis get_water_stage() which caused repeats runs_nos and an error related to incorrect numbers of positional arguments.

0.3.2

Fix bug which caused somd.rst7 files in the ensemble equilibration directories to be incorrectly numbered in some cases.
Fix bug which caused the output directory to be incorrectly replaced with “output” in some cases.
Ensure that all plotting resources get closed after analysis to avoid continually increasing memory usage.

0.3.1

Modified the SimulationRunnerIterator to ensure that the state of sub-simulation runners is saved before they are torn down. This means that setting the equilibration time at the CalcSet level will now work as expected (previously the state of the calculations was not saved, causing the analysis to fail).
Updated read_overlap_mat so that it works when “#Overlap” is not printed to a new line (but is included at the end of the previous line). This seems to happen when MBAR requires many iterations to converge.

0.3.0

Improved CalcSet so that 1) get_optimal_lam_vals can be called directly on the CalcSet object, and 2) analyse can be called directly without requiring experimental results (by specifying compare_to_exp=False).
Documentation on PME vs RF slightly improved.

0.2.1

Fixed issue https://github.com/michellab/a3fe/issues/14 where all jobs in the slurm queue were assumed to be pending or running. This caused problems when completed jobs remained in the queue.

0.2.0

Added ability to run charged compounds using the co-alchemical ion approach. Ligand net charge is detected from the input files and the config is checked to ensure that PME is used for electrostatics.
Added parameterisation tests.
Fixed bug arising from incorrect removal of “is_equilibrated” method (37de921)
Made detection of username more robust (see https://github.com/michellab/a3fe/issues/8)

0.1.3

Ensured that config options are written correctly when the template config file does not terminate in a new line. Previously, new options would be appended to the last line of the file if it did not end with n, which could happen if the user manually edited the file.

0.1.2

Fixed bug in get_slurm_file_base which caused the function to fail to read the output name if an “=” instead of a space was used to separate the argument and value.

v0.1.1

Ensured that simulation_runner objects only get pickled once they’ve been fully initialised. This avoids issues where an error occurs during initialisation and the object is pickled in an incomplete state. Addresses https://github.com/michellab/a3fe/issues/1.

v0.1.0

Initial release