Version 0.2.5

This contains numerous small changes/fixes since the v0.2.4 release, but most notably includes the introduction of water hopping as described at https://doi.org/10.26434/chemrxiv.12429464.v1.

It also includes fixes to documentation.

v0.2.4

This point release:

• adds a simple test system (charged ethylene) which can run quickly on CPU.
• Adds python 3.7 support

Version 0.2.3

• Adds the standard python .gitignore file
• Adds a .github/ which holds template files for PRs and guidelines for contributing to the code
• Adds Sphinx docs and ReadTheDocs to the repository
• Docstrings are now formatted in numpydoc style
• Code is styled in YAPF style; configured by the setup.cfg file
• Re-wrote the tests to use pytest
• Adds CodeCov to the repository
• Adds Versioneer to the repository
• Improvements to Travis-CI
  • Builds and deploys conda packages for py3.5/3.6 for both osx/linux platforms
  • Deploys the conda environment yml file onto anaconda cloud
  • Conda packages follow convention, example for version 0.2.3 on python3.5
    • blues-0.2.3-py35g{SHORTHASH}_{#UNRELEASED COMMITS} or blues-0.2.3-py35g6aaba2d_5
• Minor API changes, renaming some functions in the BLUESSimulation to adhere to conventional python name schemes
  • Move the engine.py module (which housed just one single class) into the moves.py module
  • Fix in the AlchemicalExternalLangevinIntegrator for issue #115 ref
• Make package compatible for openmm=7.2.2 and openmmtools=0.15.0
• Includes changes from #150
  • Removes some code redundancy in syncing the simulation states
  • Correctly resets the velocities before the start of each MD simulation ref
  • Corrects behavior in syncing the simulation states ref
  • Adds some checks in the freezing selection ref

Version 0.2.2

Bug fixes for OpenEye tests and restarting from the YAML; enhancements to the Logger and package installation.

• PR #144, #145 Ensures installation through setup.py installs dependencies
• PR #143 Fixes a bug when restarting through the YAML file
• PR #140 Adds the ability to stream information from the Logger module to stdout or a file
• PR #135, #139 Fixes unittests to isolate OpenEye related code, skips tests if no OE_LICENSE or OpenEye-toolskits installation is found.

Version 0.2.1

PR #125: Critical bug fix in alchemical correction term

Version 0.2.0

Release for changes from PR #117

• BLUES will now support partial configuration from a YAML file
  • Supported settings: Input/output, system parameters, simulation parameters, freezing/restraining atoms, and configuring reporters
• API Changes
  • Addition of SystemFactory class: Provides methods for freezing or restraining atoms in your system
  • Addition of SimulationFactory: Allows addition of a MonteCarloBarostat to your System, enabling NPT simulations
  • Name changes methods in the Simulation class
• Addition of reporters.py module
  • Addition of ReporterConfig class, provides functionality for generating/configuring a some recommended/customized reporters for BLUES simulations:
    • BLUESHDF5Reporter subclass from mdtraj.HDF5Reporter
      • Custom features include:
        • specifying the indices of frames to be recorded,
        • protocolWork/alchemicalLambda from the NCMC simulation
        • simulation parameters
        • exporting your conda environment
    • BLUESStateDataReporter subclass from openmm.app.StateDataReporter
      • Custom features include:
        • title field to set the string prefix that is printed with each report line.
        • enable writing to a file or file-like object for streaming the data (logging.Logger) to the terminal window,
        • specifying the indices of frames to be recorded
        • currentIter/protocolWork/alchemicalLambda from the NCMC simulation
    • NetCDF4Reporter subclass from parmed.openmm.reporters.NetCDFReporter
      • Custom features includes:
        • Add ability to properly flush the data to disc on call to report()
        • specifying the indices of frames to be recorded
        • protocolWork/alchemicalLambda from the NCMC simulation
    • parmed.openmm.reporters.RestartReporter, enabling restarts from the rst7 file

Version 0.1.3

• General enhancements to simulation logging
  • Moved initialization of the logger module into simulation.py
  • Added reporter for the NCMC simulation
  • Print out simulation information/timing
  • Lambda range to add extra propagation steps now controlled by a single parameter prop_lambda
    • Added docstrings for nprop and prop_lambda

version 0.1.2

-Introduction of a new movetype: SideChainMove

• Requires: oeommtools conda install -c openeye/label/Orion oeommtools -c omnia
• Dropped Python 2.7 support
• Fixes for the logger functionality

Version 0.1.1

• Adding extra propagation steps between lambda 0.2 -> 0.8 in the alchemical integrator
  • User parameter nprop = 5 (Default).
• Zeroing masses of atoms outside the 5.0A distance (Default) of the ligand in the alchemical system.
  • Distance for freezing atoms is now a user parameter with freeze_distance
  • Added option to set selection center freeze_center Default: 'LIG'
  • Added option to include solvent freeze_solvent Default: 'HOH,NA,CL'
• Switch print statements for logging module
• Made input parameters for functions in the Simulation class explicit
  • Set default values but can pass the input parameters from the opt dictionary
• New parameter write_move for the acceptRejectNCMC function
  • Setting write_move = True will write the accepted move to a PDB file.
• Changed runNCMC() to run().
  • Avoids potential confusion with function simulateNCMC()
• Simplified NCMC trajectory reporter
  • Replacedwrite_ncmc parameter with ncmc_traj
  • If a string is specified for (i.e. ncmc_traj = ncmc_output, it will write every NCMC step to a DCD file named ncmc_output.dcd
• Added a function _getSimulationInfo() that will print out the simulation time information

Version 0.1.0

This release:

• Adds refactored move proposal scheme
  • Allows for mixing of move types using CombinationMove
  • Allows for choosing between a set of moves with different probabilities
• Adds SmartDartMove` move class, which performs Smart Darting moves
• Adds Monte Carlo Functionality
• Switches alchemical VV integrator with an alchemical Langevin splitting integrator
• Longrange Dispersion corrections for the alchemical integrator is turned off
  • This greatly increases alchemical integration speeds (see choderalab/openmmtools#218)