def formalCharge(molecule):
"""Compute the formal charge on a molecule. This function requires that
the molecule has explicit hydrogen atoms.
Parameters
----------
molecule : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`
A molecule object.
Returns
-------
formal_charge : :class:`Charge <BioSimSpace.Types.Charge>`
The total formal charge on the molecule.
"""
if type(molecule) is not _Molecule:
raise TypeError(
"'molecule' must be of type 'BioSimSpace._SireWrappers.Molecule'")
from rdkit import Chem as _Chem
# Create a temporary working directory.
tmp_dir = _tempfile.TemporaryDirectory()
work_dir = tmp_dir.name
# Zero the total formal charge.
formal_charge = 0
# Stdout/stderr redirection doesn't work from within Jupyter.
if _is_notebook:
# Run in the working directory.
with _Utils.cd(work_dir):
# Save the molecule to a PDB file.
_IO.saveMolecules("tmp", molecule, "PDB")
# Read the ligand PDB into an RDKit molecule.
mol = _Chem.MolFromPDBFile("tmp.pdb")
# Compute the formal charge.
formal_charge = _Chem.rdmolops.GetFormalCharge(mol)
else:
# Run in the working directory and redirect stderr from RDKit.
with _Utils.cd(work_dir), _Utils.stderr_redirected():
# Save the molecule to a PDB file.
_IO.saveMolecules("tmp", molecule, "PDB")
# Read the ligand PDB into an RDKit molecule.
mol = _Chem.MolFromPDBFile("tmp.pdb")
# Compute the formal charge.
formal_charge = _Chem.rdmolops.GetFormalCharge(mol)
return formal_charge * _electron_charge
def start(self):
"""Start the SOMD process.
Returns
-------
process : :class:`Process.Somd <BioSimSpace.Process.Somd>`
A handle to the running process.
"""
# The process is currently queued.
if self.isQueued():
return
# Process is already running.
if self._process is not None:
if self._process.isRunning():
return
# Clear any existing output.
self._clear_output()
# Run the process in the working directory.
with _Utils.cd(self._work_dir):
# Create the arguments string list.
args = self.getArgStringList()
# Write the command-line process to a README.txt file.
with open("README.txt", "w") as f:
# Set the command-line string.
self._command = "%s " % self._exe + self.getArgString()
# Write the command to file.
f.write("# SOMD was run with the following command:\n")
f.write("%s\n" % self._command)
# Start the timer.
self._timer = _timeit.default_timer()
# Start the simulation.
self._process = _SireBase.Process.run(self._exe, args,
"%s.out" % self._name,
"%s.out" % self._name)
# SOMD uses the stdout stream for all output.
with open(_os.path.basename(self._stderr_file), "w") as f:
f.write(
"All output has been redirected to the stdout stream!\n")
return self
def formalCharge(molecule):
"""Compute the formal charge on a molecule. This function requires that
the molecule has explicit hydrogen atoms.
Parameters
----------
molecule : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`
A molecule object.
Returns
-------
formal_charge : :class:`Charge <BioSimSpace.Types.Charge>`
The total formal charge on the molecule.
"""
if type(molecule) is not _Molecule:
raise TypeError("'molecule' must be of type 'BioSimSpace._SireWrappers.Molecule'")
from rdkit import Chem as _Chem
from rdkit import RDLogger as _RDLogger
# Disable RDKit warnings.
_RDLogger.DisableLog('rdApp.*')
# Create a temporary working directory.
tmp_dir = _tempfile.TemporaryDirectory()
work_dir = tmp_dir.name
# Zero the total formal charge.
formal_charge = 0
# Run in the working directory.
with _Utils.cd(work_dir):
# Save the molecule to a PDB file.
_IO.saveMolecules("tmp", molecule, "PDB")
# Read the ligand PDB into an RDKit molecule.
mol = _Chem.MolFromPDBFile("tmp.pdb")
# Compute the formal charge.
formal_charge = _Chem.rdmolops.GetFormalCharge(mol)
return formal_charge * _electron_charge
def matchAtoms(molecule0,
molecule1,
scoring_function="rmsd_align",
matches=1,
return_scores=False,
prematch={},
timeout=5 * _Units.Time.second,
property_map0={},
property_map1={}):
"""Find mappings between atom indices in molecule0 to those in molecule1.
Molecules are aligned using a Maximum Common Substructure (MCS) search.
When requesting more than one match, the mappings will be sorted using
a scoring function and returned in order of best to worst score. (Note
that, depending on the scoring function the "best" score may have the
lowest value.)
Parameters
----------
molecule0 : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`
The molecule of interest.
molecule1 : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`
The reference molecule.
scoring_function : str
The scoring function used to match atoms. Available options are:
- "rmsd"
Calculate the root mean squared distance between the
coordinates of atoms in molecule0 to those that they
map to in molecule1.
- "rmsd_align"
Align molecule0 to molecule1 based on the mapping before
computing the above RMSD score.
- "rmsd_flex_align"
Flexibly align molecule0 to molecule1 based on the mapping
before computing the above RMSD score. (Requires the
'fkcombu'. package: http://strcomp.protein.osaka-u.ac.jp/kcombu)
matches : int
The maximum number of matches to return. (Sorted in order of score).
return_scores : bool
Whether to return a list containing the scores for each mapping.
prematch : dict
A dictionary of atom mappings that must be included in the match.
timeout : BioSimSpace.Types.Time
The timeout for the maximum common substructure search.
property_map0 : dict
A dictionary that maps "properties" in molecule0 to their user
defined values. This allows the user to refer to properties
with their own naming scheme, e.g. { "charge" : "my-charge" }
property_map1 : dict
A dictionary that maps "properties" in molecule1 to their user
defined values.
Returns
-------
matches : dict, [dict], ([dict], list)
The best atom mapping, a list containing a user specified number of
the best mappings ranked by their score, or a tuple containing the
list of best mappings and a list of the corresponding scores.
Examples
--------
Find the best maximum common substructure mapping between two molecules.
>>> import BioSimSpace as BSS
>>> mapping = BSS.Align.matchAtoms(molecule0, molecule1)
Find the 5 best mappings.
>>> import BioSimSpace as BSS
>>> mappings = BSS.Align.matchAtoms(molecule0, molecule1, matches=5)
Find the 5 best mappings along with their ranking scores.
>>> import BioSimSpace as BSS
>>> mappings, scores = BSS.Align.matchAtoms(molecule0, molecule1, matches=5, return_scores=True)
Find the 5 best mappings along with their ranking scores. Score
by flexibly aligning molecule0 to molecule1 based on each mapping
and computing the root mean squared displacement of the matched
atoms.
>>> import BioSimSpace as BSS
>>> mappings, scores = BSS.Align.matchAtoms(molecule0, molecule1, matches=5, return_scores=True, scoring_function="rmsd_flex_align")
Find the best mapping that contains a prematch (this is a dictionary mapping
atom indices in molecule0 to those in molecule1).
>>> import BioSimSpace as BSS
>>> mapping = BSS.Align.matchAtoms(molecule0, molecule1, prematch={0 : 10, 3 : 7})
"""
# A list of supported scoring functions.
scoring_functions = ["RMSD", "RMSDALIGN", "RMSDFLEXALIGN"]
# Validate input.
if type(molecule0) is not _Molecule:
raise TypeError(
"'molecule0' must be of type 'BioSimSpace._SireWrappers.Molecule'")
if type(molecule1) is not _Molecule:
raise TypeError(
"'molecule1' must be of type 'BioSimSpace._SireWrappers.Molecule'")
if type(scoring_function) is not str:
raise TypeError("'scoring_function' must be of type 'str'")
else:
# Strip underscores and whitespace, then convert to upper case.
_scoring_function = scoring_function.replace("_", "").upper()
_scoring_function = _scoring_function.replace(" ", "").upper()
if not _scoring_function in scoring_functions:
raise ValueError(
"Unsupported scoring function '%s'. Options are: %s" %
(scoring_function, scoring_functions))
if _scoring_function == "RMSDFLEXALIGN" and _fkcombu_exe is None:
raise _MissingSoftwareError(
"'rmsd_flex_align' option requires the 'fkcombu' program: "
"http://strcomp.protein.osaka-u.ac.jp/kcombu")
if type(matches) is not int:
raise TypeError("'matches' must be of type 'int'")
else:
if matches < 0:
raise ValueError("'matches' must be positive!")
if type(return_scores) is not bool:
raise TypeError("'return_matches' must be of type 'bool'")
if type(prematch) is not dict:
raise TypeError("'prematch' must be of type 'dict'")
else:
_validate_mapping(molecule0, molecule1, prematch, "prematch")
if type(timeout) is not _Units.Time._Time:
raise TypeError("'timeout' must be of type 'BioSimSpace.Types.Time'")
if type(property_map0) is not dict:
raise TypeError("'property_map0' must be of type 'dict'")
if type(property_map1) is not dict:
raise TypeError("'property_map1' must be of type 'dict'")
# Extract the Sire molecule from each BioSimSpace molecule.
mol0 = molecule0._getSireObject()
mol1 = molecule1._getSireObject()
# Convert the timeout to seconds and take the magnitude as an integer.
timeout = int(timeout.seconds().magnitude())
# Create a temporary working directory.
tmp_dir = _tempfile.TemporaryDirectory()
work_dir = tmp_dir.name
# Use RDKkit to find the maximum common substructure.
try:
# Run inside a temporary directory.
with _Utils.cd(work_dir):
# Write both molecules to PDB files.
_IO.saveMolecules("tmp0",
molecule0,
"PDB",
property_map=property_map0)
_IO.saveMolecules("tmp1",
molecule1,
"PDB",
property_map=property_map1)
# Load the molecules with RDKit.
# Note that the C++ function overloading seems to be broken, so we
# need to pass all arguments by position, rather than keyword.
# The arguments are: "filename", "sanitize", "removeHs", "flavor"
mols = [
_Chem.MolFromPDBFile("tmp0.pdb", False, False, 0),
_Chem.MolFromPDBFile("tmp1.pdb", False, False, 0)
]
# Generate the MCS match.
mcs = _rdFMCS.FindMCS(mols,
atomCompare=_rdFMCS.AtomCompare.CompareAny,
bondCompare=_rdFMCS.BondCompare.CompareAny,
completeRingsOnly=True,
ringMatchesRingOnly=True,
matchChiralTag=False,
matchValences=False,
maximizeBonds=False,
timeout=timeout)
# Get the common substructure as a SMARTS string.
mcs_smarts = _Chem.MolFromSmarts(mcs.smartsString)
except:
raise RuntimeError("RDKIT MCS mapping failed!")
# Score the mappings and return them in sorted order (best to worst).
mappings, scores = _score_rdkit_mappings(mol0, mol1, mols[0], mols[1],
mcs_smarts, prematch,
_scoring_function, property_map0,
property_map1)
# Sometimes RDKit fails to generate a mapping that includes the prematch.
# If so, then try generating a mapping using the MCS routine from Sire.
if len(mappings) == 1 and mappings[0] == prematch:
# Convert timeout to a Sire Unit.
timeout = timeout * _SireUnits.second
# Regular match. Include light atoms, but don't allow matches between heavy
# and light atoms.
m0 = mol0.evaluate().findMCSmatches(
mol1, _SireMol.AtomResultMatcher(_to_sire_mapping(prematch)),
timeout, True, property_map0, property_map1, 6, False)
# Include light atoms, and allow matches between heavy and light atoms.
# This captures mappings such as O --> H in methane to methanol.
m1 = mol0.evaluate().findMCSmatches(
mol1, _SireMol.AtomResultMatcher(_to_sire_mapping(prematch)),
timeout, True, property_map0, property_map1, 0, False)
# Take the mapping with the larger number of matches.
if len(m1) > 0:
if len(m0) > 0:
if len(m1[0]) > len(m0[0]):
mappings = m1
else:
mappings = m0
else:
mappings = m1
else:
mappings = m0
# Score the mappings and return them in sorted order (best to worst).
mappings, scores = _score_sire_mappings(mol0, mol1, mappings, prematch,
_scoring_function,
property_map0, property_map1)
if matches == 1:
if return_scores:
return (mappings[0], scores[0])
else:
return mappings[0]
else:
# Return a list of matches from best to worst.
if return_scores:
return (mappings[0:matches], scores[0:matches])
# Return a tuple containing the list of matches from best to
# worst along with the list of scores.
else:
return mappings[0:matches]
def flexAlign(molecule0,
molecule1,
mapping=None,
fkcombu_exe=None,
property_map0={},
property_map1={}):
"""Flexibly align atoms in molecule0 to those in molecule1 using the
mapping between matched atom indices.
Parameters
----------
molecule0 : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`
The molecule to align.
molecule1 : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`
The reference molecule.
mapping : dict
A dictionary mapping atoms in molecule0 to those in molecule1.
fkcombu_exe : str
Path to the fkcombu executable. If None is passed, then BioSimSpace
will attempt to find fkcombu by searching your PATH.
property_map0 : dict
A dictionary that maps "properties" in molecule0 to their user
defined values. This allows the user to refer to properties
with their own naming scheme, e.g. { "charge" : "my-charge" }
property_map1 : dict
A dictionary that maps "properties" in molecule1 to their user
defined values.
Returns
-------
molecule : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`
The aligned molecule.
Examples
--------
Align molecule0 to molecule1 based on a precomputed mapping.
>>> import BioSimSpace as BSS
>>> molecule0 = BSS.Align.flexAlign(molecule0, molecule1, mapping)
Align molecule0 to molecule1. Since no mapping is passed one will be
autogenerated using :class:`matchAtoms <BioSimSpace.Align.matchAtoms>`
with default options.
>>> import BioSimSpace as BSS
>>> molecule0 = BSS.Align.flexAlign(molecule0, molecule1)
"""
# Check that we found fkcombu in the PATH.
if fkcombu_exe is None:
if _fkcombu_exe is None:
raise _MissingSoftwareError(
"'BioSimSpace.Align.flexAlign' requires the 'fkcombu' program: "
"http://strcomp.protein.osaka-u.ac.jp/kcombu")
else:
fkcombu_exe = _fkcombu_exe
# Check that the user supplied executable exists.
else:
if not _os.path.isfile(fkcombu_exe):
raise IOError("'fkcombu' executable doesn't exist: '%s'" %
fkcombu_exe)
if type(molecule0) is not _Molecule:
raise TypeError(
"'molecule0' must be of type 'BioSimSpace._SireWrappers.Molecule'")
if type(molecule1) is not _Molecule:
raise TypeError(
"'molecule1' must be of type 'BioSimSpace._SireWrappers.Molecule'")
if type(property_map0) is not dict:
raise TypeError("'property_map0' must be of type 'dict'")
if type(property_map1) is not dict:
raise TypeError("'property_map1' must be of type 'dict'")
# The user has passed an atom mapping.
if mapping is not None:
if type(mapping) is not dict:
raise TypeError("'mapping' must be of type 'dict'.")
else:
_validate_mapping(molecule0, molecule1, mapping, "mapping")
# Get the best match atom mapping.
else:
mapping = matchAtoms(molecule0,
molecule1,
property_map0=property_map0,
property_map1=property_map1)
# Convert the mapping to AtomIdx key:value pairs.
sire_mapping = _to_sire_mapping(mapping)
# Create a temporary working directory.
tmp_dir = _tempfile.TemporaryDirectory()
work_dir = tmp_dir.name
# Execute in the working directory.
with _Utils.cd(work_dir):
# Write the two molecules to PDB files.
_IO.saveMolecules("molecule0",
molecule0,
"PDB",
property_map=property_map0)
_IO.saveMolecules("molecule1",
molecule1,
"PDB",
property_map=property_map1)
# Write the mapping to text. (Increment indices by one).
with open("mapping.txt", "w") as file:
for idx0, idx1 in sire_mapping.items():
file.write("%d %d\n" % (idx0.value() + 1, idx1.value() + 1))
# Create the fkcombu command string.
command = "%s -T molecule0.pdb -R molecule1.pdb -alg F -iam mapping.txt -opdbT aligned.pdb" % fkcombu_exe
# Run the command as a subprocess.
proc = _subprocess.run(command,
shell=True,
stdout=_subprocess.PIPE,
stderr=_subprocess.PIPE)
# Check that the output file exists.
if not _os.path.isfile("aligned.pdb"):
raise _AlignmentError(
"Failed to align molecules based on mapping: %r" %
mapping) from None
# Load the aligned molecule.
aligned = _IO.readMolecules("aligned.pdb")[0]
# Get the "coordinates" property for molecule0.
prop = property_map0.get("coordinates", "coordinates")
# Copy the coordinates back into the original molecule.
molecule0._sire_object = molecule0._sire_object.edit() \
.setProperty(prop, aligned._sire_object.property("coordinates")).commit()
# Return the aligned molecule.
return _Molecule(molecule0)
def _solvate(molecule,
box,
shell,
model,
num_point,
ion_conc,
is_neutral,
work_dir=None,
property_map={}):
"""Internal function to add solvent using 'gmx solvate'.
Parameters
----------
molecule : :class:`Molecule <BioSimSpace._SireWrappers.Molecule>`, \
:class:`System <BioSimSpace._SireWrappers.System>`
A molecule, or system of molecules.
box : [:class:`Length <BioSimSpace.Types.Length>`]
A list containing the box size in each dimension.
shell : :class:`Length` <BioSimSpace.Types.Length>`
Thickness of the water shell around the solute.
model : str
The name of the water model.
num_point : int
The number of atoms in the water model.
ion_conc : float
The ion concentration in (mol per litre).
is_neutral : bool
Whether to neutralise the system.
work_dir : str
The working directory for the process.
property_map : dict
A dictionary that maps system "properties" to their user defined
values. This allows the user to refer to properties with their
own naming scheme, e.g. { "charge" : "my-charge" }
Returns
-------
system : :class:`System <BioSimSpace._SireWrappers.System>`
The solvated system.
"""
if molecule is not None:
# Store the centre of the molecule.
center = molecule._getAABox(property_map).center()
# Work out the vector from the centre of the molecule to the centre of the
# water box, converting the distance in each direction to Angstroms.
vec = []
for x, y in zip(box, center):
vec.append(0.5 * x.angstroms().magnitude() - y)
# Translate the molecule. This allows us to create a water box
# around the molecule.
molecule.translate(vec, property_map)
if type(molecule) is _System:
# Reformat all of the water molecules so that they match the
# expected GROMACS topology template.
waters = _SireIO.setGromacsWater(
molecule._sire_object.search("water"), model)
# Convert to a BioSimSpace molecules container.
waters = _Molecules(waters.toMolecules())
# Remove the old water molecules then add those with the updated topology.
molecule.removeWaterMolecules()
molecule.addMolecules(waters)
# Create a temporary working directory and store the directory name.
if work_dir is None:
tmp_dir = _tempfile.TemporaryDirectory()
work_dir = tmp_dir.name
# Run the solvation in the working directory.
with _Utils.cd(work_dir):
# Create the gmx command.
if num_point == 3:
mod = "spc216"
else:
mod = model
command = "%s solvate -cs %s" % (_gmx_exe, mod)
if molecule is not None:
# Write the molecule/system to a GRO files.
_IO.saveMolecules("input", molecule, "gro87")
_os.rename("input.gro87", "input.gro")
# Update the command.
command += " -cp input.gro"
# Add the box information.
if box is not None:
command += " -box %f %f %f" % (box[0].nanometers().magnitude(),
box[1].nanometers().magnitude(),
box[2].nanometers().magnitude())
# Add the shell information.
if shell is not None:
command += " -shell %f" % shell.nanometers().magnitude()
# Just add box information.
else:
command += " -box %f %f %f" % (box[0].nanometers().magnitude(),
box[1].nanometers().magnitude(),
box[2].nanometers().magnitude())
# Add the output file.
command += " -o output.gro"
with open("README.txt", "w") as file:
# Write the command to file.
file.write("# gmx solvate was run with the following command:\n")
file.write("%s\n" % command)
# Create files for stdout/stderr.
stdout = open("solvate.out", "w")
stderr = open("solvate.err", "w")
# Run gmx solvate as a subprocess.
proc = _subprocess.run(command,
shell=True,
stdout=stdout,
stderr=stderr)
stdout.close()
stderr.close()
# gmx doesn't return sensible error codes, so we need to check that
# the expected output was generated.
if not _os.path.isfile("output.gro"):
raise RuntimeError("'gmx solvate failed to generate output!")
# Extract the water lines from the GRO file.
water_lines = []
with open("output.gro", "r") as file:
for line in file:
if _re.search("SOL", line):
# Store the SOL atom record.
water_lines.append(line)
# Add any box information. This is the last line in the GRO file.
water_lines.append(line)
# Write a GRO file that contains only the water atoms.
if len(water_lines) - 1 > 0:
with open("water.gro", "w") as file:
file.write("BioSimSpace %s water box\n" % model.upper())
file.write("%d\n" % (len(water_lines) - 1))
for line in water_lines:
file.write("%s" % line)
else:
raise ValueError(
"No water molecules were generated. Try increasing "
"the 'box' size or 'shell' thickness.")
# Create a TOP file for the water model. By default we use the Amber03
# force field to generate a dummy topology for the water model.
with open("water_ions.top", "w") as file:
file.write("#define FLEXIBLE 1\n\n")
file.write("; Include AmberO3 force field\n")
file.write('#include "amber03.ff/forcefield.itp"\n\n')
file.write("; Include %s water topology\n" % model.upper())
file.write('#include "amber03.ff/%s.itp"\n\n' % model)
file.write("; Include ions\n")
file.write('#include "amber03.ff/ions.itp"\n\n')
file.write("[ system ] \n")
file.write("BioSimSpace %s water box\n\n" % model.upper())
file.write("[ molecules ] \n")
file.write(";molecule name nr.\n")
file.write("SOL %d\n" %
((len(water_lines) - 1) / num_point))
# Load the water box.
water = _IO.readMolecules(["water.gro", "water_ions.top"])
# Create a new system by adding the water to the original molecule.
if molecule is not None:
# Translate the molecule and water back to the original position.
vec = [-x for x in vec]
molecule.translate(vec, property_map)
water.translate(vec)
if type(molecule) is _System:
# Extract the non-water molecules from the original system.
non_waters = _Molecules(
molecule.search("not water")._sire_object.toMolecules())
# Create a system by adding these to the water molecules from
# gmx solvate, which will include the original waters.
system = non_waters.toSystem() + water
else:
system = molecule.toSystem() + water
# Add all of the water box properties to the new system.
for prop in water._sire_object.propertyKeys():
prop = property_map.get(prop, prop)
# Add the space property from the water system.
system._sire_object.setProperty(
prop, water._sire_object.property(prop))
else:
system = water
# Now we add ions to the system and neutralise the charge.
if ion_conc > 0 or is_neutral:
# Write the molecule + water system to file.
_IO.saveMolecules("solvated", system, "gro87")
_IO.saveMolecules("solvated", system, "grotop")
_os.rename("solvated.gro87", "solvated.gro")
_os.rename("solvated.grotop", "solvated.top")
# First write an mdp file.
with open("ions.mdp", "w") as file:
file.write("; Neighbour searching\n")
file.write("cutoff-scheme = Verlet\n")
file.write("rlist = 1.1\n")
file.write("pbc = xyz\n")
file.write("verlet-buffer-tolerance = -1\n")
file.write("\n; Electrostatics\n")
file.write("coulombtype = cut-off\n")
file.write("\n; VdW\n")
file.write("rvdw = 1.0\n")
# Create the grompp command.
command = "%s grompp -f ions.mdp -po ions.out.mdp -c solvated.gro -p solvated.top -o ions.tpr" % _gmx_exe
with open("README.txt", "a") as file:
# Write the command to file.
file.write(
"\n# gmx grompp was run with the following command:\n")
file.write("%s\n" % command)
# Create files for stdout/stderr.
stdout = open("grommp.out", "w")
stderr = open("grommp.err", "w")
# Run grompp as a subprocess.
proc = _subprocess.run(command,
shell=True,
stdout=stdout,
stderr=stderr)
stdout.close()
stderr.close()
# Flag whether to break out of the ion adding stage.
is_break = False
# Check for the tpr output file.
if not _os.path.isfile("ions.tpr"):
if shell is None:
raise RuntimeError(
"'gmx grommp' failed to generate output! "
"Perhaps your box is too small?")
else:
is_break = True
_warnings.warn(
"Unable to achieve target ion concentration, try using "
"'box' option instead of 'shell'.")
# Only continue if grommp was successful. This allows us to skip the remainder
# of the code if the ion addition failed when the 'shell' option was chosen, i.e.
# because the estimated simulation box was too small.
if not is_break:
is_break = False
# The ion concentration is unset.
if ion_conc == 0:
# Get the current molecular charge.
charge = system.charge()
# Round to the nearest integer value.
charge = round(charge.magnitude())
# Create the genion command.
command = "echo SOL | %s genion -s ions.tpr -o solvated_ions.gro -p solvated.top -neutral" % _gmx_exe
# Add enough counter ions to neutralise the charge.
if charge > 0:
command += " -nn %d" % abs(charge)
else:
command += " -np %d" % abs(charge)
else:
# Create the genion command.
command = "echo SOL | %s genion -s ions.tpr -o solvated_ions.gro -p solvated.top -%s -conc %f" \
% (_gmx_exe, "neutral" if is_neutral else "noneutral", ion_conc)
with open("README.txt", "a") as file:
# Write the command to file.
file.write(
"\n# gmx genion was run with the following command:\n")
file.write("%s\n" % command)
# Create files for stdout/stderr.
stdout = open("genion.out", "w")
stderr = open("genion.err", "w")
# Run genion as a subprocess.
proc = _subprocess.run(command,
shell=True,
stdout=stdout,
stderr=stderr)
stdout.close()
stderr.close()
# Check for the output GRO file.
if not _os.path.isfile("solvated_ions.gro"):
if shell is None:
raise RuntimeError(
"'gmx genion' failed to add ions! Perhaps your box is too small?"
)
else:
is_break = True
_warnings.warn(
"Unable to achieve target ion concentration, try using "
"'box' option instead of 'shell'.")
if not is_break:
# Counters for the number of SOL, NA, and CL atoms.
num_sol = 0
num_na = 0
num_cl = 0
# We now need to loop through the GRO file to extract the lines
# corresponding to water or ion atoms.
water_ion_lines = []
with open("solvated_ions.gro", "r") as file:
for line in file:
# This is a Sodium atom.
if _re.search("NA", line):
water_ion_lines.append(line)
num_na += 1
# This is a Chlorine atom.
if _re.search("CL", line):
water_ion_lines.append(line)
num_cl += 1
# This is a water atom.
elif _re.search("SOL", line):
water_ion_lines.append(line)
num_sol += 1
# Add any box information. This is the last line in the GRO file.
water_ion_lines.append(line)
# Write a GRO file that contains only the water and ion atoms.
if len(water_ion_lines) - 1 > 0:
with open("water_ions.gro", "w") as file:
file.write("BioSimSpace %s water box\n" %
model.upper())
file.write("%d\n" % (len(water_ion_lines) - 1))
for line in water_ion_lines:
file.write("%s" % line)
# Ions have been added. Update the TOP file fo the water model
# with the new atom counts.
if num_na > 0 or num_cl > 0:
with open("water_ions.top", "w") as file:
file.write("#define FLEXIBLE 1\n\n")
file.write("; Include AmberO3 force field\n")
file.write(
'#include "amber03.ff/forcefield.itp"\n\n')
file.write("; Include %s water topology\n" %
model.upper())
file.write('#include "amber03.ff/%s.itp"\n\n' %
model)
file.write("; Include ions\n")
file.write('#include "amber03.ff/ions.itp"\n\n')
file.write("[ system ] \n")
file.write("BioSimSpace %s water box\n\n" %
model.upper())
file.write("[ molecules ] \n")
file.write(";molecule name nr.\n")
file.write("SOL %d\n" %
(num_sol / num_point))
if num_na > 0:
file.write("NA %d\n" % num_na)
if num_cl > 0:
file.write("CL %d\n" % num_cl)
# Load the water/ion box.
water_ions = _IO.readMolecules(
["water_ions.gro", "water_ions.top"])
# Create a new system by adding the water to the original molecule.
if molecule is not None:
if type(molecule) is _System:
# Extract the non-water molecules from the original system.
non_waters = _Molecules(
molecule.search(
"not water")._sire_object.toMolecules())
# Create a system by adding these to the water and ion
# molecules from gmx solvate, which will include the
# original waters.
system = non_waters.toSystem() + water_ions
else:
system = molecule.toSystem() + water_ions
# Add all of the water molecules' properties to the new system.
for prop in water_ions._sire_object.propertyKeys():
prop = property_map.get(prop, prop)
# Add the space property from the water system.
system._sire_object.setProperty(
prop, water_ions._sire_object.property(prop))
else:
system = water_ions
# Store the name of the water model as a system property.
system._sire_object.setProperty("water_model", _SireBase.wrap(model))
return system
def getFreeEnergy(self, index=None, stride=None, kt=None):
"""Get the current free energy estimate.
Parameters
----------
index : int
The index of the collective variable. If None, then all variables
will be considered.
stride : int
The stride for integrating the free energy. This can be used to
check for convergence.
kt : BioSimSpace.Types.Energy
The temperature in energy units for intergrating out variables.
free_energies : [:class:`Type <BioSimSpace.Types>`, ...], \
[[:class:`Type <BioSimSpace.Types>`, :class:`Type <BioSimSpace.Types>`, ...], ...]
The free energy estimate for the chosen collective variables.
"""
if index is not None:
if type(index) is not int:
raise TypeError("'index' must be of type 'int'")
if index > self._num_colvar - 1 or index < -self._num_colvar:
raise IndexError("'index' must be in range -%d to %d" %
(self._num_colvar, self._num_colvar - 1))
if stride is not None:
if type(stride) is not int:
raise TypeError("'stride' must be of type 'int'")
if stride < 0:
raise ValueError("'stride' must be >= 0")
if kt is not None:
if type(kt) is not _Types.Energy:
raise TypeError(
"'kt' must be of type 'BioSimSpace.Type.Energy'")
# Convert to kt and get the magnitude.
kt = kt.kt().magnitude()
if kt <= 0:
raise ValueError("'kt' must have magnitude > 0")
else:
if index is not None:
raise ValueError(
"You must specify 'kt' when making a dimensionality reduction."
)
# Create the command string.
command = "%s sum_hills --hills HILLS --mintozero" % self._exe
# Append additional arguments.
if index is not None:
command += " --idw %s" % self._colvar_name[index]
command += " --kt %s" % kt
if stride is not None:
command += " --stride %s" % stride
# Initialise a list to hold the free energy estimates.
free_energies = []
# Move to the working directory.
with _Utils.cd(self._work_dir):
# Run the sum_hills command as a background process.
proc = _subprocess.run(command,
shell=True,
stdout=_subprocess.PIPE,
stderr=_subprocess.PIPE)
if proc.returncode != 0:
raise RuntimeError("Failed to generate free energy estimate.\n"
"Error: %s" % proc.stderr.decode("utf-8"))
# Get a sorted list of all the fes*.dat files.
fes_files = _glob.glob("fes*.dat")
fes_files.sort()
# Process each of the files.
for fes in fes_files:
# Create a list to store the free energy estimate for this file.
free_energy = []
if index is None:
# Create lists for each the collective variable data point.
for x in range(0, self._num_colvar):
free_energy.append([])
else:
free_energy.append([])
# Create a list to store the free energy.
free_energy.append([])
# Read the file.
with open(fes, "r") as file:
# Loop over all lines in the file.
for line in file:
# Ignore comments and blank lines.
if line[0] != "#":
# Extract the data.
# This is: colvar1, colvar2, ..., fes
data = [float(x) for x in line.split()]
# The line contains data.
if len(data) > 0:
# Store data for each of the collective variables.
if index is None:
for x in range(0, self._num_colvar):
name = self._colvar_name[x]
free_energy[x].append(
data[x] * self._colvar_unit[name])
free_energy[self._num_colvar].append(
data[self._num_colvar] *
_Units.Energy.kj_per_mol)
else:
name = self._colvar_name[0]
free_energy[0].append(
data[0] * self._colvar_unit[name])
free_energy[1].append(
data[1] * _Units.Energy.kj_per_mol)
if len(fes_files) == 1:
free_energies = free_energy
else:
free_energies.append(tuple(free_energy))
# Remove the file.
_os.remove(fes)
return tuple(free_energies)