def printer(tpl,conformer,atom,count):
#template = '{0:9}{1:7}{2:5} {3:7}\n'
template = '{0:9s}{1:6s}{2:>5s}{3:+012.8f}\n'
element = oechem.OEGetAtomicSymbol(atom.GetAtomicNum()).upper()
tpl.write(template.format("RADIUS",conformer.name, \
atom.GetName(),vdw_dict[element]))
return
def PrintForces(mol, forces):
for atom in mol.GetAtoms():
oechem.OEThrow.Info(
"%6d %8s %8.2f %8.2f %8.2f" %
(atom.GetIdx(), oechem.OEGetAtomicSymbol(
atom.GetAtomicNum()), forces[atom.GetIdx()],
forces[atom.GetIdx() + 1], forces[atom.GetIdx() + 2]))
def make_ptl_mol(oemol):
"""Builds a QCPortal Molecule from an OpenEye molecule"""
coords = oemol.GetCoords()
symbols_list = [
oechem.OEGetAtomicSymbol(atom.GetAtomicNum())
for atom in mol.GetAtoms()
]
#convert to bohr
print(coords)
for key, item in coords.items():
coords[key] = (item[0] * 1.88973, item[1] * 1.88973, item[2] * 1.88973)
coord_list = [c for atom in mol.GetAtoms() for c in coords[atom.GetIdx()]]
conn_list = np.array(
[[bond.GetBgnIdx(),
bond.GetEndIdx(),
bond.GetOrder()] for bond in mol.GetBonds()])
ptl_mol = ptl.Molecule.from_data({
'geometry': coord_list,
'symbols': symbols_list,
'connectivity': conn_list
})
return ptl_mol
def atom_symbols(self):
""" return array of atomic symbols """
atom_sym = []
for at in self._mol.GetAtoms():
atom_sym.append(oechem.OEGetAtomicSymbol(at.GetAtomicNum()))
return atom_sym
def GetMolDetails(mol, label):
"""
Parameters
----------
mol: single OEChem conformer with coordinates
label: string - name of the molecule. Can be an empty string.
Returns
-------
optinfo: string - Turbomole input file for autoDefine.py
xinfo: string - XYZ format coordinates to feed into Turbomole's x2t
"""
optinfo = ("$title %s" % label)
optinfo += ("\n$charge %d" % oechem.OENetCharge(mol))
optinfo += ("\n$end")
xinfo = ("%d\n" % mol.NumAtoms())
xyz = oechem.OEFloatArray(3)
# get coordinates of each atom
for atom in mol.GetAtoms():
mol.GetCoords(atom, xyz)
xinfo+=( '\n %s %10.4f %10.4f %10.4f' \
%(oechem.OEGetAtomicSymbol(atom.GetAtomicNum()),
xyz[0], xyz[1], xyz[2]) )
return optinfo, xinfo
def to_mapped_xyz(molecule,
atom_map=None,
conformer=None,
xyz_format=True,
filename=None):
"""
Generate xyz coordinates for molecule in the order given by the atom_map. atom_map is a dictionary that maps the
tag on the SMILES to the atom idex in OEMol.
Parameters
----------
molecule: OEMol with conformers
atom_map: dict
maps tag in SMILES to atom index
conformer: int
Which conformer to write xyz file for. If None, write out all conformers. Default is None
xyz_format: bool
If True, will write out number of atoms and molecule name. If false, will only write out elements and coordinates
filename: str
Name of file to save to. If None, only returns a string.
Returns
-------
str: elements and xyz coordinates (in angstroms) in order of tagged SMILES
"""
if not atom_map and not cmiles.utils.has_atom_map(molecule):
raise ValueError(
"If molecule does not have atom map, you must provide an atom map")
if not has_conformer(molecule, check_two_dimension=True):
raise ValueError("Molecule must have conformers")
xyz = ""
for k, mol in enumerate(molecule.GetConfs()):
if k == conformer or conformer is None:
if xyz_format:
xyz += "{}\n".format(mol.GetMaxAtomIdx())
xyz += "{}\n".format(mol.GetTitle())
coords = oechem.OEFloatArray(mol.GetMaxAtomIdx() * 3)
mol.GetCoords(coords)
if k != 0 and not xyz_format:
xyz += "*"
for mapping in range(1, molecule.NumAtoms() + 1):
if not atom_map:
atom = molecule.GetAtom(oechem.OEHasMapIdx(mapping))
idx = atom.GetIdx()
else:
idx = atom_map[mapping]
atom = mol.GetAtom(oechem.OEHasAtomIdx(idx))
syb = oechem.OEGetAtomicSymbol(atom.GetAtomicNum())
xyz += " {} {:05.3f} {:05.3f} {:05.3f}\n".format(
syb, coords[idx * 3], coords[idx * 3 + 1],
coords[idx * 3 + 2])
if filename:
file = open("{}.xyz".format(filename), 'w')
file.write(xyz)
file.close()
else:
return xyz
def doAlignSs(self, unique=True, maxMatches=20):
"""Test the SS comparison between current reference and fit molecules -
Return list of corresponding atoms on success or an empty list otherwise.
"""
atomMapL = []
fitAtomUnMappedL = []
#
nAtomsRef = self.__refmol.NumAtoms()
nAtomsFit = self.__fitmol.NumAtoms()
# -------
oechem.OEAddExplicitHydrogens(self.__refmol)
oechem.OEAddExplicitHydrogens(self.__fitmol)
fitAtD = {}
#
for at in self.__fitmol.GetAtoms():
nAtL = at.GetAtoms()
neighbors = [nAt.GetName() for nAt in nAtL]
atType = oechem.OEGetAtomicSymbol(at.GetAtomicNum())
xyzL = oechem.OEFloatArray(3)
self.__fitmol.GetCoords(at, xyzL)
fitAtD[at.GetIdx()] = AlignAtomUnMapped(
fitId=self.__fitId,
fitAtIdx=at.GetIdx(),
fitAtName=at.GetName(),
fitAtType=atType,
fitAtNo=at.GetAtomicNum(),
fitAtFormalCharge=at.GetFormalCharge(),
x=xyzL[0],
y=xyzL[1],
z=xyzL[2],
fitNeighbors=neighbors,
)
#
logger.debug("nAtomsRef %d nAtomsFit %d", nAtomsRef, nAtomsFit)
#
# --------
self.__setupSubStructure(self.__refmol)
self.__ss.SetMaxMatches(maxMatches)
miter = self.__ss.Match(self.__fitmol, unique)
if miter.IsValid():
match = miter.Target()
for mAt in match.GetAtoms():
atomMapL.append(
AlignAtomMap(
refId=self.__refId,
refAtIdx=mAt.pattern.GetIdx(),
refAtNo=mAt.pattern.GetAtomicNum(),
refAtName=mAt.pattern.GetName(),
fitId=self.__fitId,
fitAtIdx=mAt.target.GetIdx(),
fitAtNo=mAt.target.GetAtomicNum(),
fitAtName=mAt.target.GetName(),
)
)
fitAtD.pop(mAt.target.GetIdx())
logger.debug("fitAtD %r", fitAtD)
fitAtomUnMappedL = list(fitAtD.values())
return (nAtomsRef, self.__refFD, nAtomsFit, self.__fitFD, atomMapL, fitAtomUnMappedL)
def doAlignMcss(self, unique=True, minFrac=1.0, useExhaustive=True):
"""Test the MCSS comparison between current reference and fit molecules -
Return list of corresponding atoms on success or an empty list otherwise.
"""
atomMapL = []
fitAtomUnMappedL = []
#
nAtomsRef = self.__refmol.NumAtoms()
nAtomsFit = self.__fitmol.NumAtoms()
fitAtD = {}
for at in self.__fitmol.GetAtoms():
nAtL = at.GetAtoms()
neighbors = [nAt.GetName() for nAt in nAtL]
atType = oechem.OEGetAtomicSymbol(at.GetAtomicNum())
xyzL = oechem.OEFloatArray(3)
self.__fitmol.GetCoords(at, xyzL)
fitAtD[at.GetIdx()] = AlignAtomUnMapped(
fitId=self.__fitId,
fitAtIdx=at.GetIdx(),
fitAtName=at.GetName(),
fitAtType=atType,
fitAtNo=at.GetAtomicNum(),
fitAtFormalCharge=at.GetFormalCharge(),
x=xyzL[0],
y=xyzL[1],
z=xyzL[2],
fitNeighbors=neighbors,
)
minAtoms = int(min(nAtomsRef, nAtomsFit) * minFrac)
# -------
self.__setupMCSS(self.__refmol, useExhaustive=useExhaustive)
self.__mcss.SetMCSFunc(oechem.OEMCSMaxAtoms())
self.__mcss.SetMinAtoms(minAtoms)
oechem.OEAddExplicitHydrogens(self.__refmol)
oechem.OEAddExplicitHydrogens(self.__fitmol)
#
# --------
miter = self.__mcss.Match(self.__fitmol, unique)
if miter.IsValid():
match = miter.Target()
for mAt in match.GetAtoms():
atomMapL.append(
AlignAtomMap(
refId=self.__refId,
refAtIdx=mAt.pattern.GetIdx(),
refAtNo=mAt.pattern.GetAtomicNum(),
refAtName=mAt.pattern.GetName(),
fitId=self.__fitId,
fitAtIdx=mAt.target.GetIdx(),
fitAtNo=mAt.target.GetAtomicNum(),
fitAtName=mAt.target.GetName(),
)
)
fitAtD.pop(mAt.target.GetIdx())
fitAtomUnMappedL = list(fitAtD.values())
return (nAtomsRef, self.__refFD, nAtomsFit, self.__fitFD, atomMapL, fitAtomUnMappedL)
def make_psi_input(mol, label, method, basisset, SPE=False, mem=None):
"""
Parameters
----------
mol: single OEChem conformer with coordinates
label: string - name of the molecule. Can be an empty string.
method: string - specification of method (see Psi4 website for options)
basisset: string - specification of basis set
SPE: boolean - False (default) for geom opt. True for single point E calcns
mem: string - specify Psi4 job memory. E.g. "2 Gb" "2000 Mb" "2000000 Kb"
Returns
-------
inputstring: string - containing contents of whole input file for this conf
"""
inputstring = ""
xyz = oechem.OEFloatArray(3)
# specify memory requirements, if defined
if mem != None:
inputstring += "memory %s\n" % mem
inputstring+=( 'molecule %s {\n' % label )
# charge and multiplicity; multiplicity hardwired to singlet (usually is)
netCharge = oechem.OENetCharge( mol)
inputstring+=( ' %s 1' % netCharge )
# get coordinates of each atom
for atom in mol.GetAtoms():
mol.GetCoords( atom, xyz)
inputstring+=( '\n %s %10.4f %10.4f %10.4f' \
%(oechem.OEGetAtomicSymbol(atom.GetAtomicNum()),
xyz[0], xyz[1], xyz[2]) )
inputstring+=( '\n units angstrom\n}')
# check if mol has a "freeze" tag
for x in oechem.OEGetSDDataPairs(mol):
if "atoms to freeze" in x.GetTag():
freeze_list = x.GetValue()
inputstring += "\n\nfreeze_list = \"\"\"\n {} xyz\n {} xyz\n {} xyz\n {} xyz\n\"\"\"".format(freeze_list[1], freeze_list[4],
freeze_list[7], freeze_list[10])
inputstring += "\nset optking frozen_cartesian $freeze_list"
inputstring += "\nset optking dynamic_level = 1\nset optking consecutive_backsteps = 2\nset optking intrafrag_step_limit = 0.1\nset optking interfrag_step_limit = 0.1"
# explicitly specify MP2 RI-auxiliary basis for Ahlrichs basis set
# http://www.psicode.org/psi4manual/master/basissets_byfamily.html
if method.lower()=='mp2' and 'def' in basisset and basisset.lower()!='def2-qzvpd':
inputstring+=('\n\nset basis %s' % (basisset))
inputstring+=('\nset df_basis_mp2 %s-ri' % (basisset))
inputstring+=('\nset freeze_core True')
else:
inputstring+=('\n\nset basis %s' % (basisset))
inputstring+=('\nset freeze_core True')
# specify command for type of calculation
if SPE is False:
inputstring+=('\noptimize(\'%s\')' % (method))
else:
inputstring+=('\nenergy(\'%s\')' % (method))
return inputstring
def printatoms(label, begin):
print(label)
for atom in begin:
print(" atom:",
atom.GetIdx(),
oechem.OEGetAtomicSymbol(atom.GetAtomicNum()),
end=" ")
print("in component",
atom.GetData(oechem.OEProperty_Component),
end=" ")
# @ <SNIPPET-RXN-MAPIDX>
print("with map index", atom.GetMapIdx())
def to_mapped_xyz(molecule,
atom_map,
conformer=None,
xyz_format=False,
filename=None):
"""
Generate xyz coordinates for molecule in the order given by the atom_map. atom_map is a dictionary that maps the
tag on the SMILES to the atom idex in OEMol.
Parameters
----------
molecule: OEMol with conformers
atom_map: dict
maps tag in SMILES to atom index
conformer: int
Which conformer to write xyz file for. If None, write out all conformers. Default is None
xyz_format: bool
If True, will write out number of atoms and molecule name. If false, will only write out elements and coordinates
filename: str
Name of file to save to. If None, only returns a string.
Returns
-------
str: elements and xyz coordinates in order of tagged SMILES
"""
xyz = ""
for k, mol in enumerate(molecule.GetConfs()):
if k == conformer or conformer is None:
if xyz_format:
xyz += "{}\n".format(mol.GetMaxAtomIdx())
xyz += "{}\n".format(mol.GetTitle())
coords = oechem.OEFloatArray(mol.GetMaxAtomIdx() * 3)
mol.GetCoords(coords)
if k != 0 and not xyz_format:
xyz += "*"
for mapping in range(1, len(atom_map) + 1):
idx = atom_map[mapping]
atom = mol.GetAtom(oechem.OEHasAtomIdx(idx))
syb = oechem.OEGetAtomicSymbol(atom.GetAtomicNum())
xyz += " {} {:05.3f} {:05.3f} {:05.3f}\n".format(
syb, coords[idx * 3], coords[idx * 3 + 1],
coords[idx * 3 + 2])
if filename:
file = open("{}.xyz".format(filename, 'w'))
file.write(xyz)
file.close()
return xyz
def oemol_to_dict(oemol, read_wbo=False):
"""
Return list of elements, partial charges and connectivity with WBOs for the bonds
Parameters
----------
oemol : oechem.OEMol
Molecule must have partial charges and Wiberg Bond Orders precalculated.
Returns
-------
mol_dict: dict
dictionary of atomic symbols, partial charges and connectivity with Wiberg Bond Orders
"""
from openeye import oechem
atomic_symbols = [
oechem.OEGetAtomicSymbol(atom.GetAtomicNum())
for atom in oemol.GetAtoms()
]
partial_charges = [atom.GetPartialCharge() for atom in oemol.GetAtoms()]
# Build connectivity with WBOs
connectivity = []
for bond in oemol.GetBonds():
a1 = bond.GetBgn().GetIdx()
a2 = bond.GetEnd().GetIdx()
if read_wbo == True:
if not 'WibergBondOrder' in bond.GetData():
raise RuntimeError('Molecule does not have Wiberg Bond Orders')
wbo = bond.GetData('WibergBondOrder')
connectivity.append([a1, a2, wbo])
else:
bond_order = bond.GetOrder()
connectivity.append([a1, a2, bond_order])
mol_dict = {
'atomic_symbols': atomic_symbols,
'partial_charges': partial_charges,
'connectivity': connectivity
}
return mol_dict
def get_map_ordered_geometry(molecule, atom_map):
"""
Generate geometry list of size 3*N in the order of the atom map
Parameters
----------
molecule: OEMol with 1 conformer or OEConf
atom_map: dict
map_idx:atom_idx
Returns
-------
symbols: list of str
list of symbols
geometry: list of ints
xyz geometry. List is length N*3
"""
if not molecule.GetDimension() == 3:
raise RuntimeError(
"Molecule must have 3D coordinates for generating a QCSchema molecule"
)
if isinstance(molecule, oechem.OEMol):
if molecule.GetMaxConfIdx() != 1:
raise Warning(
"The molecule must have at least and at most 1 conformation")
symbols = []
geometry = []
for mapping in range(1, molecule.NumAtoms() + 1):
idx = atom_map[mapping]
atom = molecule.GetAtom(oechem.OEHasAtomIdx(idx))
syb = oechem.OEGetAtomicSymbol(atom.GetAtomicNum())
symbols.append(syb)
for i in range(3):
geometry.append(molecule.GetCoords()[atom.GetIdx()][i] *
ANGSROM_2_BOHR)
return symbols, geometry
def get_psi4_mol_from_conf(conf):
"""Use an OEConfBase to construct a psi4 molecule"""
xyz = oechem.OEFloatArray(3)
mol_str = str()
total_charge = 0
for atom in conf.GetAtoms():
total_charge += atom.GetFormalCharge()
conf.GetCoords(atom, xyz)
# scaled_xyz = [c / psi4.constants.bohr2angstroms for c in xyz]
mol_str += "{} {} {} {}\n".format(
oechem.OEGetAtomicSymbol(atom.GetAtomicNum()),
# scaled_xyz[0], scaled_xyz[1], scaled_xyz[2])
xyz[0],
xyz[1],
xyz[2],
)
psi4_mol = psi4.geometry(mol_str)
psi4_mol.set_molecular_charge(total_charge)
# psi_mol.set_name('{}, conf {}'.format(conf.GetTitle(), conf.GetIdx()))
logging.debug("\n\n\nNew Psi4 molecule created with geometry: \n" +
psi4_mol.create_psi4_string_from_molecule())
return psi4_mol
# THE SAMPLE CODE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
# EXPRESS OR IMPLIED. OPENEYE DISCLAIMS ALL WARRANTIES, INCLUDING, BUT
# NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
# PARTICULAR PURPOSE AND NONINFRINGEMENT. In no event shall OpenEye be
# liable for any damages or liability in connection with the Sample Code
# or its use.
from openeye import oechem
from openeye import oegraphsim
# @ <SNIPPET-FPOVERLAP>
pmol = oechem.OEGraphMol()
oechem.OESmilesToMol(pmol, "c1cnc2c(c1)CC(CC2O)CF")
tmol = oechem.OEGraphMol()
oechem.OESmilesToMol(tmol, "c1cc2c(cc1)CC(CCl)CC2N")
fptype = oegraphsim.OEGetFPType("Tree,ver=2.0.0,size=4096,bonds=5-5,"
"atype=AtmNum|HvyDeg|EqHalo,btype=Order")
for idx, match in enumerate(oegraphsim.OEGetFPOverlap(pmol, tmol, fptype)):
ostring = "match %2d: " % (idx + 1)
for mpair in match.GetAtoms():
p = mpair.pattern
t = mpair.target
ostring += "%d%s-%d%s " % (
p.GetIdx(), oechem.OEGetAtomicSymbol(p.GetAtomicNum()), t.GetIdx(),
oechem.OEGetAtomicSymbol(t.GetAtomicNum()))
print(ostring)
# @ </SNIPPET-FPOVERLAP>
# PARTICULAR PURPOSE AND NONINFRINGEMENT. In no event shall OpenEye be
# liable for any damages or liability in connection with the Sample Code
# or its use.
# @ <SNIPPET>
from __future__ import print_function
from openeye import oechem
qfile = oechem.oemolistream("query.mol")
tfile = oechem.oemolistream("targets.sdf")
# read MDL query and initialize the substructure search
qmol = oechem.OEQMol()
oechem.OEReadMDLQueryFile(qfile, qmol)
ss = oechem.OESubSearch(qmol)
# loop over target structures
tindex = 1
for tmol in tfile.GetOEGraphMols():
oechem.OEAddExplicitHydrogens(tmol)
oechem.OEPrepareSearch(tmol, ss)
for mi in ss.Match(tmol, True):
print("hit target = %d" % tindex, end=" ")
for ai in mi.GetTargetAtoms():
print("%d%s" %
(ai.GetIdx(), oechem.OEGetAtomicSymbol(ai.GetAtomicNum())),
end=" ")
print()
tindex += 1
# @ </SNIPPET>
# THE SAMPLE CODE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
# EXPRESS OR IMPLIED. OPENEYE DISCLAIMS ALL WARRANTIES, INCLUDING, BUT
# NOT LIMITED TO, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
# PARTICULAR PURPOSE AND NONINFRINGEMENT. In no event shall OpenEye be
# liable for any damages or liability in connection with the Sample Code
# or its use.
# @ <SNIPPET>
from __future__ import print_function
from openeye import oechem
IsDonorAtom = oechem.OEMatchAtom("[!H0;#7,#8]")
mol = oechem.OEGraphMol()
oechem.OESmilesToMol(mol, "c1c(Cl)cncc1C(=O)O")
for atom in mol.GetAtoms():
atom.SetData("isdonor", IsDonorAtom(atom))
class IsDonorAtomPred(oechem.OEUnaryAtomPred):
def __call__(self, atom):
return atom.GetData("isdonor")
print("Donor atoms:", end=" ")
for atom in mol.GetAtoms(IsDonorAtomPred()):
print(atom.GetIdx(), oechem.OEGetAtomicSymbol(atom.GetAtomicNum()), end=" ")
print()
# @ </SNIPPET>
def symbol(self) -> str:
return oechem.OEGetAtomicSymbol(self._at.GetAtomicNum())
def infer_mol_from_coordinates(
coordinates,
species,
smiles_ref=None,
coordinates_unit="angstrom",
):
# local import
from openeye import oechem
from simtk import unit
from simtk.unit import Quantity
if isinstance(coordinates_unit, str):
coordinates_unit = getattr(unit, coordinates_unit)
# make sure we have the coordinates
# in the unit system
coordinates = Quantity(coordinates, coordinates_unit).value_in_unit(
unit.angstrom # to make openeye happy
)
# initialize molecule
mol = oechem.OEGraphMol()
if all(isinstance(symbol, str) for symbol in species):
[
mol.NewAtom(getattr(oechem, "OEElemNo_" + symbol))
for symbol in species
]
elif all(isinstance(symbol, int) for symbol in species):
[
mol.NewAtom(
getattr(oechem,
"OEElemNo_" + oechem.OEGetAtomicSymbol(symbol)))
for symbol in species
]
else:
raise RuntimeError(
"The species can only be all strings or all integers.")
mol.SetCoords(coordinates.reshape([-1]))
mol.SetDimension(3)
oechem.OEDetermineConnectivity(mol)
oechem.OEFindRingAtomsAndBonds(mol)
oechem.OEPerceiveBondOrders(mol)
if smiles_ref is not None:
smiles_can = oechem.OECreateCanSmiString(mol)
ims = oechem.oemolistream()
ims.SetFormat(oechem.OEFormat_SMI)
ims.openstring(smiles_ref)
mol_ref = next(ims.GetOEMols())
smiles_ref = oechem.OECreateCanSmiString(mol_ref)
assert (smiles_ref == smiles_can
), "SMILES different. Input is %s, ref is %s" % (
smiles_can,
smiles_ref,
)
from openff.toolkit.topology import Molecule
_mol = Molecule.from_openeye(mol, allow_undefined_stereo=True)
g = esp.Graph(_mol)
return g
def make_psi_input(mol, label, method, basisset, calctype='opt', mem=None):
"""
Get coordinates from input mol, and generate/format input text for
Psi4 calculation.
Parameters
----------
mol : OpenEye OEMol
OEMol with coordinates
label : string
Name of molecule with integer identifier (for conformers).
method: string
Name of the method as understood by Psi4. Example: "mp2"
basis : string
Name of the basis set as understood by Psi4. Example: "def2-sv(p)"
calctype : string
What kind of Psi4 calculation to run. Supported inputs are:
'opt' for geometry optimization,
'spe' for single point energy calculation, and
'hess' for Hessian calculation.
memory : string
How much memory each Psi4 job should take. If not specified, the
default in Psi4 is 500 Mb. Examples: "2000 MB" "1.5 GB"
http://www.psicode.org/psi4manual/master/psithoninput.html
Returns
-------
inputstring : string
Contents of Psi4 input file to be written out
"""
# check that specified calctype is valid
if calctype not in {'opt', 'spe', 'hess'}:
sys.exit("Specify a valid calculation type.")
inputstring = ""
# specify memory requirements, if defined
if mem != None:
inputstring += "memory %s\n" % mem
inputstring += ('molecule %s {\n' % label)
# charge and multiplicity; multiplicity hardwired to singlet (usually is)
netCharge = oechem.OENetCharge(mol)
inputstring += (' %s 1' % netCharge)
# get atomic symbol and coordinates of each atom
xyz = oechem.OEFloatArray(3)
for atom in mol.GetAtoms():
mol.GetCoords(atom, xyz)
inputstring+=( '\n %s %10.4f %10.4f %10.4f' \
%(oechem.OEGetAtomicSymbol(atom.GetAtomicNum()),
xyz[0], xyz[1], xyz[2]) )
inputstring += ('\n units angstrom\n}')
# check if mol has a "freeze" tag
for x in oechem.OEGetSDDataPairs(mol):
if calctype == "opt" and "atoms to freeze" in x.GetTag():
b = x.GetValue()
y = b.replace("[", "")
z = y.replace("]", "")
a = z.replace(" ", "")
freeze_list = a.split(",")
inputstring += (
"\n\nfreeze_list = \"\"\"\n {} xyz\n {} xyz\n {} "
"xyz\n {} xyz\n\"\"\"".format(freeze_list[0], freeze_list[1],
freeze_list[2], freeze_list[3]))
inputstring += "\nset optking frozen_cartesian $freeze_list"
inputstring += (
"\nset optking dynamic_level = 1\nset optking "
"consecutive_backsteps = 2\nset optking intrafrag_step_limit = "
"0.1\nset optking interfrag_step_limit = 0.1\n")
# best practices for scf calculations
# http://www.psicode.org/psi4manual/master/scf.html#recommendations
# http://www.psicode.org/psi4manual/master/dft.html#recommendations
inputstring += '\n\nset scf_type df'
inputstring += '\nset guess sad'
# explicitly specify MP2 RI-auxiliary basis for [Ahlrichs] basis set
# http://www.psicode.org/psi4manual/master/basissets_byfamily.html
# DFMP2 *should* get MP2 aux sets fine for [Pople and Dunning] sets
# http://www.psicode.org/psi4manual/master/dfmp2.html
if method.lower() == 'mp2' and 'def2' in basisset:
if basisset.lower() == 'def2-sv(p)':
inputstring += ('\nset df_basis_mp2 def2-sv_p_-ri')
elif basisset.lower() != 'def2-qzvpd': # no aux set for qzvpd 10-6-18
inputstring += ('\nset df_basis_mp2 %s-ri' % (basisset))
inputstring += ('\n\nset basis %s' % (basisset))
inputstring += ('\nset freeze_core True')
# specify command for type of calculation
if calctype == 'opt':
inputstring += ('\noptimize(\'%s\')\n\n' % (method))
elif calctype == 'spe':
inputstring += ('\nenergy(\'%s\')\n\n' % (method))
elif calctype == 'hess':
inputstring += (
'\nH, wfn = hessian(\'%s\', return_wfn=True)\nwfn.hessian().print_out()\n\n'
% (method))
return inputstring
def make_psi_json(mol, label, method, basisset, calctype='opt', mem=None):
"""
THIS FUNCTION IS A WORK IN PROGRESS.
Get coordinates from input mol, and generate/format input text for
Psi4 calculation via JSON wrapper.
Parameters
----------
mol : OpenEye OEMol
OEMol with coordinates
label : string
Name of molecule with integer identifier (for conformers).
method: string
Name of the method as understood by Psi4. Example: "mp2"
basis : string
Name of the basis set as understood by Psi4. Example: "def2-sv(p)"
calctype : string
What kind of Psi4 calculation to run. Supported inputs are:
'opt' for geometry optimization,
'spe' for single point energy calculation, and
'hess' for Hessian calculation.
memory : string
How much memory each Psi4 job should take. If not specified, the
default in Psi4 is 500 Mb. Examples: "2000 MB" "1.5 GB"
http://www.psicode.org/psi4manual/master/psithoninput.html
Returns
-------
inputstring : string
Contents of Psi4 input file to be written out
"""
# check that specified calctype is valid
if calctype not in {'opt', 'spe', 'hess'}:
sys.exit("Specify a valid calculation type.")
inputdict = {}
moldict = {}
modeldict = {}
keydict = {}
inputdict["schema_name"] = "qc_schema_input"
inputdict["schema_version"] = 1
# specify memory requirements, if defined
if mem != None:
inputdict["memory"] = mem
#TODO -- json version
# charge and multiplicity; multiplicity hardwired to singlet (usually is)
#inputdict["charge"] = oechem.OENetCharge( mol)
#inputdict["multiplicity"] = 1
# get atomic symbol and coordinates of each atom
geom_list = []
elem_list = []
xyz = oechem.OEFloatArray(3)
for atom in mol.GetAtoms():
mol.GetCoords(atom, xyz)
geom_list.append(xyz[0])
geom_list.append(xyz[1])
geom_list.append(xyz[2])
elem_list.append(oechem.OEGetAtomicSymbol(atom.GetAtomicNum()))
moldict["geometry"] = geom_list
moldict["symbols"] = elem_list
inputdict["molecule"] = moldict
#TODO -- json version
## check if mol has a "freeze" tag
#for x in oechem.OEGetSDDataPairs(mol):
# if calctype=="opt" and "atoms to freeze" in x.GetTag():
# b = x.GetValue()
# y = b.replace("[", "")
# z = y.replace("]", "")
# a = z.replace(" ", "")
# freeze_list = a.split(",")
# inputstring += ("\n\nfreeze_list = \"\"\"\n {} xyz\n {} xyz\n {} "
# "xyz\n {} xyz\n\"\"\"".format(freeze_list[0],
# freeze_list[1], freeze_list[2], freeze_list[3]))
# inputstring += "\nset optking frozen_cartesian $freeze_list"
# inputstring += ("\nset optking dynamic_level = 1\nset optking "
# "consecutive_backsteps = 2\nset optking intrafrag_step_limit = "
# "0.1\nset optking interfrag_step_limit = 0.1\n")
#TODO -- json version
## explicitly specify MP2 RI-auxiliary basis for Ahlrichs basis set
## http://www.psicode.org/psi4manual/master/basissets_byfamily.html
## DFMP2 *should* get MP2 aux sets fine for Pople/Dunning
## http://www.psicode.org/psi4manual/master/dfmp2.html
#if method.lower()=='mp2' and 'def2' in basisset:
# if basisset.lower()=='def2-sv(p)':
# inputstring+=('\nset df_basis_mp2 def2-sv_p_-ri')
# elif basisset.lower()!='def2-qzvpd': # no aux set for qzvpd 10-6-18
# inputstring+=('\nset df_basis_mp2 %s-ri' % (basisset))
modeldict["basis"] = basisset
modeldict["method"] = method
inputdict["model"] = modeldict
#inputstring+=('\nset freeze_core True')
# specify command for type of calculation
if calctype == 'opt':
# TODO
pass
elif calctype == 'spe':
inputdict["driver"] = 'energy'
elif calctype == 'hess':
inputdict["driver"] = 'hessian'
keydict["return_wfn"] = True
inputdict["keywords"] = keydict
return inputdict
def __str__(self):
return "Atom not assigned: %6d %8s" % (
self.atom.GetIdx(),
oechem.OEGetAtomicSymbol(self.atom.GetAtomicNum()))
def __makeChemCompAtomCategory(self, ccId, oeMol, writeIdealXyz=True):
"""Populate elements of chemical component definition for atoms and bonds."""
#
idCode = ccId
#
rowL = []
for ii, atom in enumerate(oeMol.GetAtoms()):
#
atRow = {}
atRow["comp_id"] = idCode
atRow["atom_id"] = atom.GetName().strip()
atRow["alt_atom_id"] = atom.GetName().strip()
atRow["type_symbol"] = oechem.OEGetAtomicSymbol(atom.GetAtomicNum())
atRow["charge"] = atom.GetFormalCharge()
(xC, yC, zC) = oeMol.GetCoords(atom)
if writeIdealXyz:
atRow["pdbx_model_Cartn_x_ideal"] = "%0.3f" % xC
atRow["pdbx_model_Cartn_y_ideal"] = "%0.3f" % yC
atRow["pdbx_model_Cartn_z_ideal"] = "%0.3f" % zC
else:
atRow["model_Cartn_x"] = "%0.3f" % xC
atRow["model_Cartn_y"] = "%0.3f" % yC
atRow["model_Cartn_z"] = "%0.3f" % zC
if atom.GetStringData("pdbx_leaving_atom_flag") in ["Y", "N"]:
atRow["pdbx_leaving_atom_flag"] = atom.GetStringData("pdbx_leaving_atom_flag")
else:
atRow["pdbx_leaving_atom_flag"] = "N"
if len(atRow["atom_id"]) > 3 or len(atRow["type_symbol"]) == 2:
atRow["pdbx_align"] = 0
else:
atRow["pdbx_align"] = 1
if atom.IsAromatic():
atRow["pdbx_aromatic_flag"] = "Y"
else:
atRow["pdbx_aromatic_flag"] = "N"
# oeSt = oechem.OEGetCIPStereo(mol, atom)
oeSt = None
# if atom.IsChiral():
cip = oechem.OEPerceiveCIPStereo(oeMol, atom)
if atom.HasStereoSpecified():
if cip == oechem.OECIPAtomStereo_S:
oeSt = "S"
if cip == oechem.OECIPAtomStereo_R:
oeSt = "R"
if cip == oechem.OECIPAtomStereo_NotStereo:
oeSt = None
if cip == oechem.OECIPAtomStereo_UnspecStereo:
oeSt = "U"
# oeSt = oechem.OEGetCIPStereo(oeMol, atom)
# oeSt = atom.GetCIPStereo()
if oeSt is not None and (len(oeSt) > 0) and (oeSt == "R" or oeSt == "S"):
atRow["pdbx_stereo_config"] = oeSt
else:
atRow["pdbx_stereo_config"] = "N"
atRow["pdbx_component_atom_id"] = atom.GetName().strip()
atRow["pdbx_component_comp_id"] = idCode
atRow["pdbx_ordinal"] = str(ii + 1)
rowL.append(atRow)
#
return rowL
'halognes': [],
'P': [],
'S': [],
'all': [],
'all_no_h': []}
missing_wbos = set()
for m in mols:
for bond in m.GetBonds():
if 'WibergBondOrder' not in bond.GetData():
missing_wbos.add(m)
continue
wbo = bond.GetData('WibergBondOrder')
wbos['all'].append(wbo)
a1 = bond.GetBgn()
a2 = bond.GetEnd()
sym1 = oechem.OEGetAtomicSymbol(a1.GetAtomicNum())
sym2 = oechem.OEGetAtomicSymbol(a2.GetAtomicNum())
if bond.IsInRing():
wbos['ring'].append(bond.GetData('WibergBondOrder'))
if sym1 == 'N' or sym2 == 'N':
wbos['N'].append(wbo)
if sym1 == 'S' or sym2 == 'S':
wbos['S'].append(wbo)
if sym1 == 'P' or sym2 == 'P':
wbos['P'].append(wbo)
if sym1 == 'O' or sym2 == 'O':
wbos['O'].append(wbo)
if (sym1 == 'C' and sym2 == 'C') and not bond.IsInRing() and not sym1 == 'H' and not sym2 == 'H':
wbos['C-C'].append(wbo)
if sym1 == 'H' or sym2 == 'H':
def get_symbols(molecule):
return [
oechem.OEGetAtomicSymbol(a.GetAtomicNum())
for a in molecule.GetAtoms()
]
def _featurize_atom(atom: oechem.OEAtomBase) -> np.ndarray:
"""Get atom feature vector
atom (oechem.OEAtomBase): OEAtomBase object
"""
results = (
Nodes.one_of_k_encoding(
oechem.OEGetAtomicSymbol(atom.GetAtomicNum()),
[
"C",
"N",
"O",
"S",
"F",
"Si",
"P",
"Cl",
"Br",
"Mg",
"Na",
"Ca",
"Fe",
"As",
"Al",
"I",
"B",
"V",
"K",
"Tl",
"Yb",
"Sb",
"Sn",
"Ag",
"Pd",
"Co",
"Se",
"Ti",
"Zn",
"H", # H?
"Li",
"Ge",
"Cu",
"Au",
"Ni",
"Cd",
"In",
"Mn",
"Zr",
"Cr",
"Pt",
"Hg",
"Pb",
"Unknown",
], unk=True
)
+ Nodes.one_of_k_encoding(
atom.GetDegree(), [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
)
+ Nodes.one_of_k_encoding(
atom.GetImplicitHCount(), [0, 1, 2, 3, 4, 5, 6], unk=True
)
+ [atom.GetFormalCharge()]
+ Nodes.one_of_k_encoding(
atom.GetHyb(),
[
oechem.OEHybridization_Unknown,
oechem.OEHybridization_sp,
oechem.OEHybridization_sp2,
oechem.OEHybridization_sp3,
oechem.OEHybridization_sp3d,
oechem.OEHybridization_sp3d2,
],
)
+ [atom.IsAromatic()]
+ Nodes.one_of_k_encoding(
atom.GetExplicitHCount(), [0, 1, 2, 3, 4], unk=True
),
)
return np.array(results)
def has_stereo_defined(molecule):
"""
Check if any stereochemistry in undefined.
Parameters
----------
molecule: OEMol
Returns
-------
bool: True if all stereo chemistry is defined.
If any stereochemsitry is undefined, raise and exception.
"""
# perceive stereochemistry
oechem.OEPerceiveChiral(molecule)
oechem.OE3DToAtomStereo(molecule)
oechem.OE3DToBondStereo(molecule)
unspec_chiral = False
unspec_db = False
problematic_atoms = list()
problematic_bonds = list()
for atom in molecule.GetAtoms():
if atom.IsChiral() and not atom.HasStereoSpecified(
oechem.OEAtomStereo_Tetrahedral):
# Check if handness is specified
v = []
for nbr in atom.GetAtoms():
v.append(nbr)
stereo = atom.GetStereo(v, oechem.OEAtomStereo_Tetrahedral)
if stereo == oechem.OEAtomStereo_Undefined:
unspec_chiral = True
problematic_atoms.append(
(atom.GetIdx(),
oechem.OEGetAtomicSymbol(atom.GetAtomicNum())))
for bond in molecule.GetBonds():
if bond.IsChiral() and not bond.HasStereoSpecified(
oechem.OEBondStereo_CisTrans):
v = []
for neigh in bond.GetBgn().GetAtoms():
if neigh != bond.GetEnd():
v.append(neigh)
break
for neigh in bond.GetEnd().GetAtoms():
if neigh != bond.GetBgn():
v.append(neigh)
break
stereo = bond.GetStereo(v, oechem.OEBondStereo_CisTrans)
if stereo == oechem.OEBondStereo_Undefined:
unspec_db = True
a1 = bond.GetBgn()
a2 = bond.GetEnd()
a1_idx = a1.GetIdx()
a2_idx = a2.GetIdx()
a1_s = oechem.OEGetAtomicSymbol(a1.GetAtomicNum())
a2_s = oechem.OEGetAtomicSymbol(a2.GetAtomicNum())
bond_order = bond.GetOrder()
problematic_bonds.append(
(a1_idx, a1_s, a2_idx, a2_s, bond_order))
if unspec_chiral or unspec_db:
warnings.warn(
"Stereochemistry is unspecified. Problematic atoms {}, problematic bonds {}, SMILES: {}"
.format(problematic_atoms, problematic_bonds,
oechem.OEMolToSmiles(molecule)))
return False
else:
return True
# PARTICULAR PURPOSE AND NONINFRINGEMENT. In no event shall OpenEye be
# liable for any damages or liability in connection with the Sample Code
# or its use.
# @ <SNIPPET>
from __future__ import print_function
from openeye import oechem
import sys
if len(sys.argv) != 2:
oechem.OEThrow.Usage("%s <input>" % sys.argv[0])
ifs = oechem.oemolistream()
if not ifs.open(sys.argv[1]):
oechem.OEThrow.Fatal("Unable to open %s" % sys.argv[1])
for mol in ifs.GetOEGraphMols():
print(mol.NumAtoms())
print(mol.GetTitle())
coords = oechem.OEFloatArray(mol.GetMaxAtomIdx() * 3)
mol.GetCoords(coords)
for atom in mol.GetAtoms():
idx = atom.GetIdx()
syb = oechem.OEGetAtomicSymbol(atom.GetAtomicNum())
print("%-3s%11.5f%11.5f%11.5f" %
(syb, coords[idx * 3], coords[idx * 3 + 1], coords[idx * 3 + 2]))
# @ </SNIPPET>
def debugTypes(self, mol):
for atom in mol.GetAtoms():
print "%6d %8s %8s" % (
atom.GetIdx(), oechem.OEGetAtomicSymbol(
atom.GetAtomicNum()), atom.GetStringData(self.pattyTag))
def __str__(self): return "%i%s" % (self._idx,
oechem.OEGetAtomicSymbol(self.atomic_number()))
def atomic_number(self): return self.atom.GetAtomicNum()
