• BREAKS API: Renamed AlchemicalFactory to AbsoluteAlchemicalFactory (#206)
• Major optimizations of ThermodynamicState (#200, #205)
  • keep in memory only a single System object per compatible state
  • fast copy/deepcopy
  • enable custom optimized serialization for multiple states
• Added readthedocs documentation (#191)
• Bugfix for serialization of context when NaN encountered (#199)
• Added tests for python 3.6 (#184)
• Added tests for integrators (#186, #187)

Major changes

• Overhaul of LangevinIntegrator and subclasses to better support nonequilibrium integrators
• Add true reaction-field support to AlchemicalFactory
• Add some alchemical test systems

Updates to openmmtools.integrators.LangevinIntegrator and friends

API-breaking changes:

• The nonequilibrium integrators are now called AlchemicalNonequilibriumLangevinIntegrator and ExternalPerturbationLangevinIntegrator, and both are subclasses of a common NonequilibriumLangevinIntegrator that provides a consistent interface to setting and getting protocol_work
• AlchemicalNonequilibriumLangevinIntegrator now has a default alchemical_functions to eliminate need for every test to treat it as a special case (#180)
• The get_protocol_work() method allows you to retrieve the protocol work from any NonequilibriumLangevinIntegrator subclass and returns a unit-bearing work. The optional dimensionless=True argument returns a dimensionless float in units of kT.
• Integrator global variables now store all energies in natural OpenMM units (kJ/mol) but the new accessor methods (see below) should b used instead of getting integrator global variables for work and heat. (#181)
• Any private methods for adding steps to the integrator have been prepended with _ to hide them from the public API.

New features

• Order of arguments for all LangevinIntegrator derivatives matches openmm.LangevinIntegrator so it can act as a drop-in replacement. (#176)
• The get_shadow_work() and get_heat() methods are now available for any LangevinIntegrator subclass, as well as the corresponding properties shadow_work and heat. The functions also support dimensionless=True. (#163)
• The shadow_work and heat properties were added to all LangevinIntegrator subclasses, returning the values of these properties (if the integrator was constructed with the appropriate measure_shadow_work=True or measure_heat=True flags) as unit-bearing quantities
• The get_protocol_work() and get_total_work() methods are now available for any NonequilibriumLangevinIntegrator, returning unit-bearing quantities unless dimensionless=True is provided in which case they return the work in implicit units of kT. get_total_work() requires the integrator to have been constructed with measure_shadow_work=True.
• The protocol_work and total_work properties were added to all NonequilibriumLangevinIntegrator subclasses, and return the unit-bearing work quantities. total_work requires the integrator to have been constructed with measure_shadow_work=True.
• The subclasses have been reworked to support any kwargs that the base classes support, and defaults have all been made consistent.
• Various reset() methods have been added to reset statistics for all LangevinIntegrator subclasses.
• All custom integrators support .pretty_format() and .pretty_print() with optional highlighting of specific step types.

Bugfixes

• Zero-step perturbations now work correctly (#177)
• AlchemicalNonequilibriumLangevinIntegrator now correctly supports multiple H steps.

Internal changes

• Adding new LangevinIntegrator step methods now uses a self._register_step_method(step_string, callback_function, supports_force_groups=False) call to simplify this process.
• Code duplication has been reduced through the use of calling base class methods whenever possible.
• run_nonequilibrium_switching() test now uses BAR to test dragging a harmonic oscillator and tests a variety of integrator splittings (["O { V R H R V } O", "O V R H R V O", "R V O H O V R", "H R V O V R H"]).
• Integrator tests use deterministic PME and mixed precision when able.

Updates to openmmtools.alchemy.AlchemicalFactory

Reaction field electrostatics now removes the shift, setting c_rf = 0.

A convenience method AlchemicalFactory.replace_reaction_field() has been added to allow fully-interacting systems to be modified to force c_rf = 0 by recoding reaction-field electrostatics as a CustomNonbondedForce

New openmmtools.testsystems classes

• AlchemicalWaterBox was added, which has the first water molecule in the system alchemically modified

This update enables resetting the work, and other statistics of the ExternalPerturbationLangevinIntegrator by setting the first_step global variable to 0.

This release includes updates to the Langevin integrators with variable splitting schemes, such as BAOAB and others.

This release fixes a Python 2 bug caused by how Python 2 and Python 3 handle differently user-defined equality operators. This made the sanity checks in alchemy.AlchemicalState to always fail when checking the compatibility of a System.

This version adds several improvements to the integrators module, it introduces a general framework to perform MCMC simulations, and it imports the functionalities of choderalab/alchemy.

• Added a LangevinSplittingIntegrator class that includes several popular integrators of Langevin dynamics as special cases.
• Added BAOABIntegrator and GeodesicBAOABIntegrator.
• New base CustomIntegrators: ThermostatedIntegrator, which offers utilities for integrators that maintain the distribution at a certain temperature, and RestorableIntegrator that enable restoring class members (e.g. temperature getter/setters) that are otherwise lost with OpenMM serialization.
• New state classes ThermodynamicState, SamplerState and CompoundThermodynamicState that can be used to store and manipulate OpenMM Systems and Contexts.
• New class ContextCache that centralizes the creation and caching of OpenMM Contexts by exploiting the mechanism of compatibility between ThermodynamicStates.
• Imported existing MCMC framework (previously in choderalab/openmmmcmc) and adapted it to the new state classes. MCMCMoves share a ContextCache to minimize the number of created Contexts during the simulation.
• Imported the alchemy module and adapted it to the new framework. AlchemicalState can be used with CompoundThermodynamicState to extend ThermodynamicState by composition.

This release extends some of the systems in openmmtools.testsystems to allow multiple CustomGBForce-based GB models supported by OpenMM's customgbforces.py functionality to be used.

• TolueneImplicit, HostGuestImplicit, and LysozymeImplicit now accept the implicitSolvent option, allowing GB models from simtk.openmm.app (such as HCT, OBC1, OBC2, GBn, and GBn2) to be optionally specified. OBC1 remains the default.
• These three classes also support any kwargs supported by AmberPrmtopFile.createSystem(). Any parameters that createSystem supports will be passed along from the constructor.
• TolueneImplicit and HostGuestImplicit also have variants like TolueneImplicitHCT, TolueneImplicitOBC1, etc. for all OpenMM-supported implicit solvent models.

Fixes a py3 issue in installation.

• Fixed a bug in GHMCIntegrator where updateContextState() was not being called
• Improvements to GHMCIntegrator for speed
• potential_new and potential_old have been added as global parameters for GHMCIntegrator

• Added a host-guest system (CB7:B2):
  • HostGuestVacuum
  • HostGuestImplicit
  • HostGuestExplicit
• GHMCIntegrator now includes protection from nans