def init(self, **kwargs):
"""
Parameters
----------
**kwargs :
Returns
-------
"""
super().init(**kwargs)
# load the json topology as an mdtraj one
mdtraj_top = json_to_mdtraj_topology(self.json_main_rep_top)
# make a traj for the initial state to use as a topology for
# visualizing the walkers
init_traj = mdj.Trajectory(
[self.init_main_rep_positions],
unitcell_lengths=[self.init_unitcell_lengths],
unitcell_angles=[self.init_unitcell_angles],
topology=mdtraj_top)
# write out the init traj as a pdb
logging.info("Writing initial state to {}".format(
self.init_state_path))
init_traj.save_pdb(self.init_state_path)
def binding_site_idxs(json_topology, coords, box_vectors, cutoff):
# convert quantities to numbers in nanometers
cutoff = cutoff.value_in_unit(unit.nanometer)
coords = coords.value_in_unit(unit.nanometer)
box_lengths, box_angles = box_vectors_to_lengths_angles(
box_vectors.value_in_unit(unit.nanometer))
# selecting ligand and protein binding site atom indices for
# resampler and boundary conditions
lig_idxs = ligand_idxs(json_topology)
prot_idxs = protein_idxs(json_topology)
# make a trajectory to compute the neighbors from
traj = mdj.Trajectory(np.array([coords]),
unitcell_lengths=[box_lengths],
unitcell_angles=[box_angles],
topology=json_to_mdtraj_topology(json_topology))
# selects protein atoms which have less than 8 A from ligand
# atoms in the crystal structure
neighbors_idxs = mdj.compute_neighbors(traj, cutoff, lig_idxs)
# selects protein atoms from neighbors list
binding_selection_idxs = np.intersect1d(neighbors_idxs, prot_idxs)
return binding_selection_idxs
def _calc_min_distance(self, walker):
"""Min-min distance for a walker.
Parameters
----------
walker : object implementing the Walker interface
Returns
-------
min_distance : float
"""
cell_lengths, cell_angles = box_vectors_to_lengths_angles(
walker.state['box_vectors'])
# convert the json topology to an mdtraj one
mdj_top = json_to_mdtraj_topology(self._topology)
# make a traj out of it so we can calculate distances through
# the periodic boundary conditions
walker_traj = mdj.Trajectory(walker.state['positions'],
topology=mdj_top,
unitcell_lengths=cell_lengths,
unitcell_angles=cell_angles)
# calculate the distances through periodic boundary conditions
# and get hte minimum distance
min_distance = np.min(
mdj.compute_distances(
walker_traj, it.product(self.ligand_idxs, self.receptor_idxs)))
return min_distance
def init(self, **kwargs):
"""
Parameters
----------
**kwargs :
Returns
-------
"""
super().init(**kwargs)
if self.init_image is not None:
image_mdj_topology = json_to_mdtraj_topology(
self.json_main_rep_top)
# initialize the initial image into the image traj
init_image_traj = mdj.Trajectory([self.init_image],
time=self.times,
topology=image_mdj_topology)
# save this as a PDB for a topology to view in VMD etc. to go
# along with the trajectory we will make
logging.info("Writing initial image to {}".format(
self.init_state_path))
init_image_traj.save_pdb(self.init_state_path)
self._top_pdb_written = True
def report(self, cycle_idx=None, new_walkers=None, **kwargs):
"""Report the current cycle's walker states as 3D molecular
structures.
Parameters
----------
cycle_idx : int
new_walkers : list of Walker objects
"""
# load the json topology as an mdtraj one
mdtraj_top = json_to_mdtraj_topology(self.json_main_rep_top)
# slice off the main_rep indices because that is all we want
# to write for these
main_rep_positions = np.array([
walker.state['positions'][self.main_rep_idxs]
for walker in new_walkers
])
# convert the box vectors
unitcell_lengths, unitcell_angles = traj_box_vectors_to_lengths_angles(
np.array([walker.state['box_vectors'] for walker in new_walkers]))
# make a trajectory from these walkers
traj = mdj.Trajectory(main_rep_positions,
unitcell_lengths=unitcell_lengths,
unitcell_angles=unitcell_angles,
topology=mdtraj_top)
# write to the file for this trajectory
traj.save_dcd(self.walker_path)
def report(self, cycle_idx=None, resampler_data=None, **kwargs):
# load the json topology as an mdtraj one
image_mdj_topology = json_to_mdtraj_topology(self.json_main_rep_top)
# collect the new images defined
new_images = []
for resampler_rec in resampler_data:
image = resampler_rec['image']
new_images.append(image)
times = np.array([cycle_idx + 1 for _ in range(len(new_images))])
# combine the new image positions and times with the old
self.image_traj_positions.extend(new_images)
self.times.extend(times)
# only save if we have an image yet
if len(self.image_traj_positions) > 0:
# make a trajectory of the new images, using the cycle_idx as the time
new_image_traj = mdj.Trajectory(self.image_traj_positions,
time=self.times,
topology=image_mdj_topology)
# if we haven't already written a topology PDB write it now
if not self._top_pdb_written:
new_image_traj[0].save_pdb(self.init_state_path)
self._top_pdb_written = True
# then the images to the trajectory file
new_image_traj.save_dcd(self.image_path)
def binding_site_idxs(json_topology,
ligand_idxs,
receptor_idxs,
coords,
box_vectors,
cutoff,
periodic=True):
"""
Parameters
----------
json_topology : str
ligand_idxs : arraylike (1,)
receptor_idxs : arraylike (1,)
coords : simtk.Quantity
box_vectors : simtk.Quantity
cutoff : float
Returns
-------
binding_site_idxs : arraylike (1,)
"""
# convert quantities to numbers in nanometers
cutoff = cutoff.value_in_unit(unit.nanometer)
coords = coords.value_in_unit(unit.nanometer)
box_lengths, box_angles = box_vectors_to_lengths_angles(
box_vectors.value_in_unit(unit.nanometer))
# make a trajectory to compute the neighbors from
traj = mdj.Trajectory(np.array([coords]),
unitcell_lengths=[box_lengths],
unitcell_angles=[box_angles],
topology=json_to_mdtraj_topology(json_topology))
neighbors_idxs = mdj.compute_neighbors(traj,
cutoff,
ligand_idxs,
periodic=periodic)[0]
# selects protein atoms from neighbors list
binding_selection_idxs = np.intersect1d(neighbors_idxs, receptor_idxs)
return binding_selection_idxs
def init(self, **kwargs):
"""Initialize the reporter at simulation time.
This will generate the initial state PDB file.
"""
super().init(**kwargs)
# load the json topology as an mdtraj one
mdtraj_top = json_to_mdtraj_topology(self.json_main_rep_top)
# make a traj for the initial state to use as a topology for
# visualizing the walkers
init_traj = mdj.Trajectory(
[self.init_main_rep_positions],
unitcell_lengths=[self.init_unitcell_lengths],
unitcell_angles=[self.init_unitcell_angles],
topology=mdtraj_top)
# write out the init traj as a pdb
logging.info("Writing initial state to {}".format(
self.init_state_path))
init_traj.save_pdb(self.init_state_path)
def binding_site_idxs(json_topology,
ligand_idxs,
receptor_idxs,
coords,
box_vectors,
cutoff,
periodic=True):
"""Parameters
----------
json_topology : str
ligand_idxs : arraylike (1,)
receptor_idxs : arraylike (1,)
coords : N x 3 arraylike of float or simtk.Quantity
If not a quantity will implicitly be treated as being in
nanometers.
box_vectors : simtk.Quantity
If not a quantity will implicitly be treated as being in
nanometers.
cutoff : float or simtk.Quantity
If not a quantity will implicitly be treated as being in
nanometers.
Returns
-------
binding_site_idxs : arraylike (1,)
"""
# if they are simtk.units convert quantities to numbers in
# nanometers
if unit.is_quantity(cutoff):
cutoff = cutoff.value_in_unit(unit.nanometer)
if unit.is_quantity(coords):
coords = coords.value_in_unit(unit.nanometer)
if unit.is_quantity(box_vectors):
box_vectors = box_vectors.value_in_unit(unit.nanometer)
box_lengths, box_angles = box_vectors_to_lengths_angles(box_vectors)
# make a trajectory to compute the neighbors from
traj = mdj.Trajectory(np.array([coords]),
unitcell_lengths=[box_lengths],
unitcell_angles=[box_angles],
topology=json_to_mdtraj_topology(json_topology))
neighbors_idxs = mdj.compute_neighbors(traj,
cutoff,
ligand_idxs,
periodic=periodic)[0]
# selects protein atoms from neighbors list
binding_selection_idxs = np.intersect1d(neighbors_idxs, receptor_idxs)
return binding_selection_idxs
def __init__(self, initial_state=None,
cutoff_distance=1.0,
topology=None,
ligand_idxs=None,
receptor_idxs=None,
periodic=True,
**kwargs):
"""Constructor for UnbindingBC class.
All the key-word arguments are necessary.
The 'initial_state' should be the initial state of your
simulation for proper non-equilibrium simulations.
Arguments
---------
initial_state : object implementing State interface
The state walkers will take on after unbinding.
cutoff_distance : float
The distance that specifies the boundary condition. When
the min-min ligand-receptor distance is less than this it
will be warped.
topology : str
A JSON string of topology.
ligand_idxs : list of int
Indices of the atoms in the topology that correspond to the ligands.
receptor_idxs : list of int
Indices of the atoms in the topology that correspond to the
receptor for the ligand.
Raises
------
AssertionError
If any of the following are not provided: initial_state, topology,
ligand_idxs, receptor_idxs
AssertionError
If the cutoff distance is not a float.
Warnings
--------
The 'initial_state' should be the initial state of your
simulation for proper non-equilibrium simulations.
Notes
-----
The topology argument is necessary due to an implementation
detail that uses mdtraj and may not be required in the future.
"""
# since the super class can handle multiple initial states we
# wrap the single initial state to a list.
super().__init__(initial_states=[initial_state],
ligand_idxs=ligand_idxs,
receptor_idxs=receptor_idxs,
**kwargs)
# test input
assert topology is not None, "Must give a reference topology"
assert type(cutoff_distance) is float
self._cutoff_distance = cutoff_distance
self._topology = topology
# convert the json topology to an mdtraj one
self._mdj_top = json_to_mdtraj_topology(self._topology)
# whether or not to use the periodic box vectors in the
# distance calculation
self._periodic = periodic
def run_sim(init_state_path, json_top_path, forcefield_paths, n_cycles,
n_steps, n_workers, **kwargs):
#### Wepy Orchestrator
# load the wepy.OpenMMState
with open(init_state_path, 'rb') as rf:
init_state = pickle.load(rf)
### Apparatus
# Runner components
# load the JSON for the topology
with open(json_top_path) as rf:
json_top_str = rf.read()
# load it with mdtraj and then convert to openmm
mdj_top = json_to_mdtraj_topology(json_top_str)
omm_topology = mdj_top.to_openmm()
# we need to use the box vectors for setting the simulation up,
# paying mind to the units
box_vectors = init_state['box_vectors'] * init_state.box_vectors_unit
# set the box to the last box size from equilibration
omm_topology.setPeriodicBoxVectors(box_vectors)
# force field parameters
force_field = omma.ForceField(*forcefield_paths)
# create a system using the topology method giving it a topology and
# the method for calculation
runner_system = force_field.createSystem(omm_topology,
nonbondedMethod=NONBONDED_METHOD,
nonbondedCutoff=NONBONDED_CUTOFF,
constraints=MD_CONSTRAINTS,
rigidWater=RIGID_WATER,
removeCMMotion=REMOVE_CM_MOTION,
hydrogenMass=HYDROGEN_MASS)
# barostat to keep pressure constant
runner_barostat = omm.MonteCarloBarostat(PRESSURE, TEMPERATURE,
VOLUME_MOVE_FREQ)
# add it to the system
runner_system.addForce(runner_barostat)
# set up for a short simulation to runner and prepare
# instantiate an integrator
runner_integrator = omm.LangevinIntegrator(TEMPERATURE,
FRICTION_COEFFICIENT, STEP_TIME)
## Runner
runner = OpenMMRunner(runner_system,
omm_topology,
runner_integrator,
platform=PLATFORM)
## Resampler
# Distance Metric
# TODO set distance metric
distance_metric = None
# TODO set resampler
resampler = None
## Boundary Conditions
# TODO optional: set the boundary conditions
bc = None
# apparatus = WepySimApparatus(runner, resampler=resampler,
# boundary_conditions=bc)
print("created apparatus")
## CONFIGURATION
# the idxs of the main representation to save in the output files,
# it is just the protein and the ligand
# TODO optional: set the main representation atom indices
main_rep_idxs = None
# REPORTERS
# list of reporter classes and partial kwargs for using in the
# orchestrator
hdf5_reporter_kwargs = {
'main_rep_idxs': main_rep_idxs,
'topology': json_top_str,
'resampler': resampler,
'boundary_conditions': bc,
# general parameters
'save_fields': SAVE_FIELDS,
'units': dict(UNITS),
'sparse_fields': dict(SPARSE_FIELDS),
'all_atoms_rep_freq': ALL_ATOMS_SAVE_FREQ
}
# get all the reporters together. Order is important since they
# will get paired with the kwargs
reporter_classes = [
WepyHDF5Reporter,
]
# collate the kwargs in the same order
reporter_kwargs = [
hdf5_reporter_kwargs,
]
# make the configuration with all these reporters and the default number of workers
configuration = Configuration(n_workers=DEFAULT_N_WORKERS,
reporter_classes=reporter_classes,
reporter_partial_kwargs=reporter_kwargs,
config_name="no-orch")
# then instantiate them
reporters = configuration._gen_reporters()
print("created configuration")
### Initial Walkers
init_walkers = [
Walker(deepcopy(init_state), INIT_WEIGHT) for _ in range(N_WALKERS)
]
print("created init walkers")
### Orchestrator
# orchestrator = Orchestrator(apparatus,
# default_init_walkers=init_walkers,
# default_configuration=configuration)
### Work Mapper
if PLATFORM in ('OpenCL', 'CUDA'):
# we use a mapper that uses GPUs
work_mapper = WorkerMapper(worker_type=OpenMMGPUWorker,
num_workers=n_workers)
if PLATFORM in ('Reference', 'CPU'):
# we just use the standard mapper
work_mapper = Mapper
### Simulation Manager
sim_manager = Manager(init_walkers,
runner=runner,
resampler=resampler,
boundary_conditions=bc,
work_mapper=work_mapper,
reporters=reporters)
### Run the simulation
steps = [n_steps for _ in range(n_cycles)]
sim_manager.run_simulation(n_cycles, steps)
def run_sim(init_state_path,
json_top_path,
forcefield_paths,
n_cycles,
n_steps,
platform,
n_workers,
lig_ff=None,
**kwargs):
# add in the ligand force fields
assert lig_ff is not None, "must give ligand forcefield"
forcefield_paths.append(lig_ff)
#### Wepy Orchestrator
# load the wepy.OpenMMState
with open(init_state_path, 'rb') as rf:
init_state = pickle.load(rf)
### Apparatus
# Runner components
# load the JSON for the topology
with open(json_top_path) as rf:
json_top_str = rf.read()
# load it with mdtraj and then convert to openmm
mdj_top = json_to_mdtraj_topology(json_top_str)
omm_topology = mdj_top.to_openmm()
# we need to use the box vectors for setting the simulation up,
# paying mind to the units
box_vectors = init_state['box_vectors'] * init_state.box_vectors_unit
positions = init_state['positions'] * init_state.positions_unit
# set the box to the last box size from equilibration
omm_topology.setPeriodicBoxVectors(box_vectors)
# force field parameters
force_field = omma.ForceField(*forcefield_paths)
# create a system using the topology method giving it a topology and
# the method for calculation
runner_system = force_field.createSystem(omm_topology,
nonbondedMethod=NONBONDED_METHOD,
nonbondedCutoff=NONBONDED_CUTOFF,
constraints=MD_CONSTRAINTS,
rigidWater=RIGID_WATER,
removeCMMotion=REMOVE_CM_MOTION,
hydrogenMass=HYDROGEN_MASS)
# barostat to keep pressure constant
runner_barostat = omm.MonteCarloBarostat(PRESSURE, TEMPERATURE,
VOLUME_MOVE_FREQ)
# add it to the system
runner_system.addForce(runner_barostat)
# set up for a short simulation to runner and prepare
# instantiate an integrator
runner_integrator = omm.LangevinIntegrator(TEMPERATURE,
FRICTION_COEFFICIENT, STEP_TIME)
## Runner
runner = OpenMMRunner(runner_system,
omm_topology,
runner_integrator,
platform=platform)
## Resampler
# Distance Metric
lig_idxs = ligand_idxs(json_top_str)
prot_idxs = protein_idxs(json_top_str)
bs_idxs = binding_site_idxs(json_top_str, positions, box_vectors, CUTOFF)
# set distance metric
distance_metric = UnbindingDistance(lig_idxs, bs_idxs, init_state)
# set resampler
resampler = WExploreResampler(distance=distance_metric,
init_state=init_state,
max_n_regions=MAX_N_REGIONS,
max_region_sizes=MAX_REGION_SIZES,
pmin=PMIN,
pmax=PMAX)
## Boundary Conditions
# optional: set the boundary conditions
bc = None
## CONFIGURATION
# the idxs of the main representation to save in the output files,
# it is just the protein and the ligand
# optional: set the main representation atom indices, set to None
# to save all the atoms in the 'positions' field
main_rep_idxs = np.concatenate((lig_idxs, prot_idxs))
# REPORTERS
# list of reporter classes and partial kwargs for using in the
# orchestrator
hdf5_reporter_kwargs = {
'main_rep_idxs': main_rep_idxs,
'topology': json_top_str,
'resampler': resampler,
'boundary_conditions': bc,
# general parameters
'save_fields': SAVE_FIELDS,
'units': dict(UNITS),
'sparse_fields': dict(SPARSE_FIELDS),
'all_atoms_rep_freq': ALL_ATOMS_SAVE_FREQ
}
# get all the reporters together. Order is important since they
# will get paired with the kwargs
reporter_classes = [
WepyHDF5Reporter,
]
# collate the kwargs in the same order
reporter_kwargs = [
hdf5_reporter_kwargs,
]
# make the configuration with all these reporters and the default
# number of workers. Don't be thrown off by this. You don't need
# this. It is just a convenient way to dynamically name the
# outputs of the reporters and parametrize the workers and worker
# mappers. This is mainly for use in the Orchestrator framework
# but it is useful here just for batch naming everything.
configuration = Configuration(n_workers=DEFAULT_N_WORKERS,
reporter_classes=reporter_classes,
reporter_partial_kwargs=reporter_kwargs,
config_name="no-orch",
mode='w')
# then instantiate the reporters from the configuration. THis
# localizes the file paths to outputs and applies the key-word
# arguments specified above.
reporters = configuration._gen_reporters()
print("created configuration")
### Initial Walkers
init_walkers = [
Walker(deepcopy(init_state), INIT_WEIGHT) for _ in range(N_WALKERS)
]
print("created init walkers")
### Work Mapper
if platform in ('OpenCL', 'CUDA'):
# we use a mapper that uses GPUs
work_mapper = WorkerMapper(worker_type=OpenMMGPUWorker,
num_workers=n_workers)
elif platform in ('CPU', ):
# for the CPU we can choose how many threads to use per walker.
worker_attributes = {'num_threads': N_CPU_THREADS}
work_mapper = WorkerMapper(worker_type=OpenMMCPUWorker,
worker_attributes=worker_attributes,
num_workers=n_workers)
elif platform in ('Reference', ):
# we just use the standard mapper for in serial
work_mapper = Mapper
### Simulation Manager
sim_manager = Manager(init_walkers,
runner=runner,
resampler=resampler,
boundary_conditions=bc,
work_mapper=work_mapper,
reporters=reporters)
### Run the simulation
steps = [n_steps for _ in range(n_cycles)]
sim_manager.run_simulation(n_cycles, steps)
