def get_sd_data(mol, tag):
try:
if oechem.OEHasSDData(mol, tag):
return oechem.OEGetSDData(mol, tag)
if oechem.OEHasSDData(mol.GetActive(), tag):
return oechem.OEGetSDData(mol.GetActive(), tag)
except AttributeError as e:
print(e)
return ""
def GetScoreToCmp(mol):
if oechem.OEHasSDData(mol, "ShapeTanimoto"):
# sort by shape tanimoto
if oechem.OEHasSDData(mol, "TanimotoCombo"):
return float(oechem.OEGetSDData(mol, "TanimotoCombo"))
return float(oechem.OEGetSDData(mol, "ShapeTanimoto"))
else:
# sort by shape tversky
if oechem.OEHasSDData(mol, "TverskyCombo"):
return float(oechem.OEGetSDData(mol, "TverskyCombo"))
return float(oechem.OEGetSDData(mol, "ShapeTversky"))
def extractXY(fname, tag):
ifs = oechem.oemolistream()
ifs.SetConfTest(oechem.OEAbsoluteConfTest())
if not ifs.open(fname):
oechem.OEThrow.Warning("Unable to open %s for reading" % fname)
xlist = []
ylist = []
for mol in ifs.GetOEMols():
for j, conf in enumerate(mol.GetConfs()):
xlist.append(int(mol.GetTitle()))
try:
ylist.append(float(oechem.OEGetSDData(conf, tag)))
except ValueError as e:
print("Missing tag data for mol {}, conf {}! Skipping.".format(
mol.GetTitle(), j))
print(e)
xlist.pop()
### convert to numpy array, take relative e, convert to kcal/mol
ylist = np.array(ylist)
ylist = ylist - ylist[0]
ylist = 627.5095 * ylist
return xlist, ylist
def main(**kwargs):
outfn = os.path.splitext(opt['infn'])[0] + '_' + opt['suffix'] + '.mol2'
success = False
### Read in .sdf file and distinguish each molecule's conformers
ifs = oechem.oemolistream()
ifs.SetConfTest(oechem.OEAbsoluteConfTest())
if not ifs.open(opt['infn']):
oechem.OEThrow.Warning("Unable to open %s for reading" % opt['infn'])
return
for mol in ifs.GetOEMols():
if mol.GetTitle() == opt['title']:
# write out all confs in this mol if no SD tag is specified
if opt['sdtag'] == "":
success = True
write_conf_mol(outfn, mol)
return
# look for the conformer in this mol with specific SD tag value
for i, conf in enumerate(mol.GetConfs()):
if oechem.OEGetSDData(conf, opt['sdtag']) == opt['value']:
success = True
write_conf_mol(outfn, conf)
if not success:
print("\n** Found no confs matching your criteria. **")
ifs.close()
def nfragments_score(molecule):
overlapping_fragments = oechem.OEGetSDData(molecule,
'overlapping_fragments')
if overlapping_fragments == '':
return 0
else:
return -len(overlapping_fragments.split(','))
def SDF2CSV(ifs, csv):
taglist = []
# read through once to find all unique tags
for mol in ifs.GetOEGraphMols():
for dp in oechem.OEGetSDDataPairs(mol):
if dp.GetTag() not in taglist:
taglist.append(dp.GetTag())
ifs.rewind()
# print out column labels
header = "Title"
for tag in taglist:
header += ",%s" % tag
header += '\n'
csv.write(header)
# build csv file
for mol in ifs.GetOEGraphMols():
line = [mol.GetTitle()]
for tag in taglist:
if oechem.OEHasSDData(mol, tag):
value = oechem.OEGetSDData(mol, tag)
else:
value = ''
line.append(',')
line.append(value)
csv.write(''.join(line))
csv.write('\n')
def generate_restricted_conformers_star(args):
core_smarts_list = [
'C1(CCOc2ccccc12)C(=O)Nc1cncc2ccccc12', # benzopyran-linker-isoquinoline
'C(=O)Nc1cncc2ccccc12', # linker-isoquinoline
]
mols = [
generate_restricted_conformers(*args, core_smarts=core_smarts)
for core_smarts in core_smarts_list
]
# Sprint 5 special: Select unproblematic enantiomers by energy
if mols[1] is None:
return mols[0]
try:
from openeye import oechem
energies = [
float(oechem.OEGetSDData(mol, 'MMFF_internal_energy'))
for mol in mols
]
if energies[0] > energies[1] + 100:
return mols[1]
else:
return mols[0]
except Exception as e:
print(e)
return None
def get_result(self, query_id):
cur_rank = list()
url = self.args.url + "/queries/{}/".format(query_id)
response = None
tries = 0
while response == None or data["status"]["job"] != "COMPLETED":
time.sleep(60 * tries)
tries += 1
response = requests.get(url)
data = response.json()
results_url = data["results"]
results_data = requests.get(self.args.url + results_url)
with tempfile.NamedTemporaryFile(suffix='.oeb',
mode='wb',
delete=False) as temp:
temp.write(results_data.content)
temp.flush()
with oechem.oemolistream(temp.name) as results:
for mol in results.GetOEGraphMols():
cur_rank.append(
(oechem.OEMolToSmiles(mol), mol.GetTitle(),
float(oechem.OEGetSDData(mol, 'TanimotoCombo')),
self.baitset[0], False))
os.remove(temp.name)
return cur_rank
def get_results(self):
self.cur_scores = {}
for query_id in self.query_id_list:
url = self.args.url + "/queries/{}/".format(query_id)
response = None
tries = 0
while response == None or data["status"]["job"] != "COMPLETED":
tries += 1
time.sleep(tries)
response = requests.get(url)
data = response.json()
results_url = data["results"]
results_data = requests.get(self.args.url + results_url)
with tempfile.NamedTemporaryFile(suffix='.oeb',
mode='wb',
delete=False) as temp:
temp.write(results_data.content)
temp.flush()
with oechem.oemolistream(temp.name) as results:
for mol in results.GetOEGraphMols():
if self.dataset_infos[1][
mol.GetTitle()] not in self.baitset[1]:
tanimoto_combo = float(
oechem.OEGetSDData(mol, "TanimotoCombo"))
if mol.GetTitle() in self.cur_scores.keys():
if self.cur_scores[
mol.GetTitle()][0] < tanimoto_combo:
self.cur_scores[mol.GetTitle()] = (
tanimoto_combo, mol.CreateCopy())
else:
self.cur_scores[mol.GetTitle()] = (
tanimoto_combo, mol.CreateCopy())
os.remove(temp.name)
def score(mol):
text = oechem.OEGetSDData(mol, 'docking_score')
try:
score = float(text)
return score
except Exception as e:
return 0
def RenderData(image, mol, tags):
from openeye import oechem
from openeye import oedepict
data = []
for tag in tags:
value = "N/A"
if oechem.OEHasSDData(mol, tag):
value = oechem.OEGetSDData(mol, tag)
data.append((tag, value))
nrdata = len(data)
tableopts = oedepict.OEImageTableOptions(
nrdata, 2, oedepict.OEImageTableStyle_LightBlue)
tableopts.SetColumnWidths([10, 20])
tableopts.SetMargins(2.0)
tableopts.SetHeader(False)
tableopts.SetStubColumn(True)
table = oedepict.OEImageTable(image, tableopts)
for row, (tag, value) in enumerate(data):
cell = table.GetCell(row + 1, 1)
table.DrawText(cell, tag + ":")
cell = table.GetBodyCell(row + 1, 1)
table.DrawText(cell, value)
def generate_restricted_conformers_star(args):
core_smarts_list = [
'C1(CCOc2ccccc12)C(=O)Nc1cncc2ccccc12', # benzopyran-linker-isoquinoline
'CNc1cncc2ccccc12', # linker-isoquinoline
'c1cncc2ccccc12', # isoquinoline
]
# Build with all core_smarts options
mols = [ generate_restricted_conformers(*args, core_smarts=core_smarts) for core_smarts in core_smarts_list ]
# Prune unsuccessful builds
mols = [ mol for mol in mols if mol!=None ]
# DEBUG
if len(mols) == 0:
# No options available
logging.warning('No core_smarts variations succeeded.')
return None
if len(mols) == 1:
# Return the only choice
return mols[0]
try:
# Prefer the second option if strain energy of first option is poor
from openeye import oechem
energies = [ float(oechem.OEGetSDData(mol, 'MMFF_internal_energy')) for mol in mols ]
if energies[0] > energies[1] + 100:
return mols[1]
else:
return mols[0]
except Exception as e:
logging.warning(e)
return None
def min_ffxml(mol, ffxml):
# make copy of the input mol
oe_mol = oechem.OEGraphMol(mol)
try:
# create openforcefield molecule ==> prone to triggering Exception
off_mol = Molecule.from_openeye(oe_mol)
# load in force field
ff = ForceField(ffxml)
# create components for OpenMM system
topology = Topology.from_molecules(molecules=[off_mol])
# create openmm system ==> prone to triggering Exception
#system = ff.create_openmm_system(topology, charge_from_molecules=[off_mol])
system = ff.create_openmm_system(topology)
except Exception as e:
smilabel = oechem.OEGetSDData(oe_mol, "SMILES QCArchive")
print( ' >>> openforcefield failed to create OpenMM system: '
f"'{oe_mol.GetTitle()}' '{smilabel}'")
print(f"{e}\n")
return
print(" >>> successful OpenMM system creation for openforcefield "
f"mol \"{oe_mol.GetTitle()}\"")
def generate_training_input(mol_file):
'''
:param mol_file: str
:return: pd.DataFrame
'''
ifs = oechem.oemolistream(mol_file)
training_data = []
for mol in ifs.GetOEGraphMols():
energy = float(oechem.OEGetSDData(mol, ENERGY_KEY))
sf_elements = get_sf_elements(mol)
dihe_inchi = get_dihedral_inchi_key(mol)
data = [dihe_inchi, energy]
data.extend(sf_elements)
training_data.append(data)
ifs.close()
columns = [INCHI_KEY, ENERGY_KEY]
num_sf_elements = len(training_data[0]) - 2
sf_columns = ['sf_%d' % (i + 1) for i in range(num_sf_elements)]
columns.extend(sf_columns)
df = pd.DataFrame(training_data, columns=columns)
# calculate relative energy for each profile
grouped = df.loc[:, [INCHI_KEY, ENERGY_KEY]].groupby(INCHI_KEY)
df2 = grouped.transform(lambda x: x - x.min())
df[ENERGY_KEY] = df2[ENERGY_KEY]
return df
def has_ic50(mol):
"""Return True if this molecule has fluorescence IC50 data"""
from openeye import oechem
try:
pIC50 = oechem.OEGetSDData(mol, 'f_avg_pIC50')
pIC50 = float(pIC50)
return True
except Exception as e:
return False
def main(argv=[__name__]):
itf = oechem.OEInterface(InterfaceData, argv)
if not (itf.HasDouble("-min") or itf.HasDouble("-max")):
oechem.OEThrow.Fatal("Please set a filter value with -min or -max")
ifs = oechem.oemolistream()
if not ifs.open(itf.GetString("-i")):
oechem.OEThrow.Fatal("Unable to open %s for reading" %
itf.GetString("-i"))
if not oechem.OEIsSDDataFormat(ifs.GetFormat()):
oechem.OEThrow.Fatal(
"Only works for input file formats that support SD data (sdf,oeb,csv)"
)
ofs = oechem.oemolostream()
if not ofs.open(itf.GetString("-o")):
oechem.OEThrow.Fatal("Unable to open %s for writing" %
itf.GetString("-i"))
if not oechem.OEIsSDDataFormat(ofs.GetFormat()):
oechem.OEThrow.Fatal(
"Only works for output file formats that support SD data \
(sdf,oeb,csv)")
tag = itf.GetString("-tag")
minval = float("-inf")
if itf.HasDouble("-min"):
minval = itf.GetDouble("-min")
maxval = float("inf")
if itf.HasDouble("-max"):
maxval = itf.GetDouble("-max")
for mol in ifs.GetOEGraphMols():
if not oechem.OEHasSDData(mol, tag):
oechem.OEThrow.Warning("Unable to find %s tag on %s" %
(tag, mol.GetTitle()))
continue
value = oechem.OEGetSDData(mol, tag)
try:
tagvalue = float(value)
except ValueError:
oechem.OEThrow.Warning("Failed to convert (%s) to a number in %s" %
(value, mol.GetTitle()))
continue
if tagvalue < minval:
continue
if tagvalue > maxval:
continue
oechem.OEWriteMolecule(ofs, mol)
def Rename(ifs, ofs, fieldname):
for mol in ifs.GetOEGraphMols():
if oechem.OEHasSDData(mol, fieldname):
mol.SetTitle(oechem.OEGetSDData(mol, fieldname))
else:
title = mol.GetTitle()
oechem.OEThrow.Warning(
"Renaming of molecule %s failed; no field %s" %
(title, fieldname))
oechem.OEWriteMolecule(ofs, mol)
def dock_molecule(molecule, default_receptor='x0387'):
"""
Dock the specified molecules, writing out to specified file
Parameters
----------
molecule : OEMol
The molecule to dock
default_receptor : str, optional, default='0387'
The default receptor to dock to
Returns
-------
all_docked_molecules : list of OEMol
All docked molecules
"""
import os
import oechem
# Make a copy of the molecule
molecule = oechem.OEMol(molecule)
# Extract list of corresponding receptor(s)
fragments = list()
fragments = oechem.OEGetSDData(molecule, "fragments").split(',')
fragments = [
fragment for fragment in fragments
if os.path.exists(f'../receptors/Mpro-{fragment}-receptor.oeb.gz')
]
if len(fragments) == 0:
fragments = [default_receptor]
# Dock them
all_docked_molecules = list()
for fragment in fragments:
molecule_to_dock = oechem.OEMol(molecule)
import os
receptor_filename = os.path.join(
f'../receptors/Mpro-{fragment}-receptor.oeb.gz')
oechem.OESetSDData(molecule_to_dock, "fragments", fragment)
# Enumerate reasonable protomers/tautomers
from openeye import oequacpac
protomer = oechem.OEMol()
protomers = [
oechem.OEMol(protomer) for protomer in
oequacpac.OEGetReasonableProtomers(molecule_to_dock)
]
docked_molecules = dock_molecules_to_receptor(receptor_filename,
protomers)
all_docked_molecules += docked_molecules
return all_docked_molecules
def is_active(molecule):
"""
"""
from openeye import oechem
value = oechem.OEGetSDData(molecule, 'IC50 (uM)')
try:
IC50 = float(value)
if (IC50 < 99.0):
return True
except:
pass
return False
def GetUrlSDF2SMI(url, fout, ntries=20, poll_wait=10):
'''Get PubChem SDF.GZ, convert to SMILES using the SD tag PUBCHEM_OPENEYE_CAN_SMILES
or PUBCHEM_OPENEYE_ISO_SMILES, and PUBCHEM_COMPOUND_CID or PUBCHEM_COMPOUND_SID for name.'''
import openeye.oechem as oechem
def HandleOEErrors(oeerrs, nowarn):
errstr = oeerrs.str()
for line in errstr.split('\n'):
if not line.rstrip(): continue
if re.search('Warning', line, re.I) and nowarn: continue
sys.stderr.write("%s\n" % line)
oeerrs.clear()
fout_tmp = tempfile.NamedTemporaryFile(prefix='pubchem_ftp_',
suffix='.sdf.gz',
delete=False)
GetUrl(url, fout_tmp, ntries, poll_wait)
fpath_tmp = fout_tmp.name
logging.debug('fpath_tmp = %s' % fpath_tmp)
fout_tmp.close()
ims = oechem.oemolistream(fpath_tmp)
ims.SetFormat(oechem.OEFormat_SDF)
ims.Setgz(True)
mol = oechem.OEGraphMol()
nbytes = 0
oeerrs = oechem.oeosstream()
oechem.OEThrow.SetOutputStream(oeerrs)
while oechem.OEReadMolecule(ims, mol):
cid = oechem.OEGetSDData(mol, 'PUBCHEM_COMPOUND_CID')
cansmi = oechem.OEGetSDData(mol, 'PUBCHEM_OPENEYE_CAN_SMILES')
isosmi = oechem.OEGetSDData(mol, 'PUBCHEM_OPENEYE_ISO_SMILES')
buff = ("%s %s\n" % (isosmi, cid))
fout.write(buff)
nbytes += len(buff)
HandleOEErrors(oeerrs, True)
os.remove(fpath_tmp)
return nbytes
def read_molecules(file_path, verbose=True):
"""
Read molecules from an OpenEye-supported file.
Parameters
----------
file_path : str
Filename from which molecules are to be read (e.g. mol2, sdf)
Returns
-------
molecules : list of OEMol
List of molecules read from file
"""
warnings.warn(DEPRECATION_WARNING_TEXT, PendingDeprecationWarning)
from openeye import oechem
from openforcefield.utils import get_data_file_path
if not os.path.exists(file_path):
built_in = get_data_file_path(f"molecules/{file_path}")
if not os.path.exists(built_in):
raise Exception(f"File '{file_path}' not found.")
file_path = built_in
if verbose: print(f"Loading molecules from '{file_path}'...")
start_time = time.time()
molecules = list()
input_molstream = oechem.oemolistream(file_path)
flavor = oechem.OEIFlavor_Generic_Default | oechem.OEIFlavor_MOL2_Default | oechem.OEIFlavor_MOL2_Forcefield
input_molstream.SetFlavor(oechem.OEFormat_MOL2, flavor)
molecule = oechem.OECreateOEGraphMol()
while oechem.OEReadMolecule(input_molstream, molecule):
# If molecule has no title, try getting SD 'name' tag
if molecule.GetTitle() == '':
name = oechem.OEGetSDData(molecule, 'name').strip()
molecule.SetTitle(name)
# Append to list.
molecule_copy = oechem.OEMol(molecule)
molecules.append(molecule_copy)
input_molstream.close()
if verbose: print(f"{len(molecules)} molecules read")
end_time = time.time()
elapsed_time = end_time - start_time
if verbose: print(f"{elapsed_time:.3f} s elapsed")
return molecules
def plotSDF(infile, tag, figname='lineplot.png'):
"""
Parameters
----------
"""
### Read in .sdf file and distinguish each molecule's conformers
ifs = oechem.oemolistream()
ifs.SetConfTest(oechem.OEAbsoluteConfTest())
if not ifs.open(infile):
oechem.OEThrow.Warning("Unable to open %s for reading" % infile)
return
xlist = []
ylist = []
for mol in ifs.GetOEMols():
print(mol.GetTitle(), mol.NumConfs())
for j, conf in enumerate(mol.GetConfs()):
try:
ylist.append(float(oechem.OEGetSDData(conf, tag)))
xlist.append(int(mol.GetTitle()))
except ValueError as err:
pass # mols not converged may not have tag
### convert to numpy array, take relative e, convert to kcal/mol
ylist = np.array(ylist)
ylist = ylist - ylist[0]
ylist = 627.5095 * ylist
### Plot.
xlabel = 'conformation number'
ylabel = "Relative energy (kcal/mol)"
fig = plt.figure()
# ax = fig.gca()
# ax.set_xticks(np.arange(-1,RefNumConfs+1,2))
plt.ylabel(ylabel, fontsize=14)
plt.xlabel(xlabel, fontsize=14)
plt.scatter(xlist, ylist)
# plt.plot(xlist, ylist)
plt.ylim(-1, 16)
# plt.xticks(range(RefNumConfs),xlabs,fontsize=12)
# plt.yticks(fontsize=12)
plt.grid()
plt.savefig(figname, bbox_inches='tight')
plt.show()
def find_string_tag(infile):
# read input file
ifs = oechem.oemolistream()
ifs.SetConfTest(oechem.OEAbsCanonicalConfTest())
if not ifs.open(infile):
oechem.OEThrow.Warning("Unable to open input file for reading")
# loop through and evaluate tags
for mol in ifs.GetOEMols():
for conf in mol.GetConfs():
mytag = oechem.OEGetSDData(conf, 'SMILES QCArchive')
count1 = mytag.count('S')
count2 = mytag.count('P')
count3 = mytag.count('C#N')
count4 = mytag.count('N/N')
print(f"{conf.GetTitle()}\t{count1}\t{count2}\t{count3}\t{count4}")
def min_mmff94x(mol, ofs, mmff94s=False):
"""
Minimize the mol with MMFF94 or MMFF94S force field.
Parameters
----------
mol : OpenEye single-conformer molecule
ofs : OpenEye output filestream
mmff94s : Boolean
True to minimize with MMFF94S
"""
# make copy of the input mol
oe_mol = oechem.OEGraphMol(mol)
# set general energy options along with the run type specification
optSzybki = oeszybki.OESzybkiOptions()
optSzybki.SetSolventModel(oeszybki.OESolventModel_NoSolv)
optSzybki.SetOptimizerType(oeszybki.OEOptType_BFGS)
# minimize with input charges not mmff94(s) charges
# https://docs.eyesopen.com/toolkits/python/szybkitk/examples.html#optimization-of-all-conformers-of-a-ligand
optSzybki.GetSolventOptions().SetChargeEngine(oequacpac.OEChargeEngineNoOp())
# set the particular force field
if mmff94s:
sdlabel = "MMFF94S"
optSzybki.SetForceFieldType(oeszybki.OEForceFieldType_MMFF94S)
else:
sdlabel = "MMFF94"
optSzybki.SetForceFieldType(oeszybki.OEForceFieldType_MMFF94)
# generate minimization engine
szOpt = oeszybki.OESzybki(optSzybki)
# make object to hold szybki results
szResults = oeszybki.OESzybkiResults()
# perform minimization
if not szOpt(oe_mol, szResults):
smilabel = oechem.OEGetSDData(oe_mol, "SMILES QCArchive")
print( ' >>> MMFF94x minimization failed for %s\n' % smilabel )
energy = szResults.GetTotalEnergy()
# save geometry, save energy as tag, write mol to file
oechem.OESetSDData(oe_mol, f"Energy {sdlabel}", str(energy))
oechem.OEWriteConstMolecule(ofs, oe_mol)
def min_ffxml(mol, ofs, ffxml):
"""
Minimize the mol with force field input from FFXML file.
Parameters
----------
mol : OpenEye single-conformer molecule
ofs : OpenEye output filestream
ffxml : string
name of FFXML file
"""
# make copy of the input mol
oe_mol = oechem.OEGraphMol(mol)
try:
# create openforcefield molecule ==> prone to triggering Exception
off_mol = Molecule.from_openeye(oe_mol)
# load in force field
ff = ForceField(ffxml)
# create components for OpenMM system
topology = Topology.from_molecules(molecules=[off_mol])
# create openmm system ==> prone to triggering Exception
#system = ff.create_openmm_system(topology, charge_from_molecules=[off_mol])
system = ff.create_openmm_system(topology)
except Exception as e:
smilabel = oechem.OEGetSDData(oe_mol, "SMILES QCArchive")
print( ' >>> openforcefield failed to create OpenMM system: '
f'{oe_mol.GetTitle()} {smilabel}: {e}')
return
positions = structure.extractPositionsFromOEMol(oe_mol)
# minimize structure with ffxml
newpos, energy = run_openmm(topology, system, positions)
# save geometry, save energy as tag, write mol to file
oe_mol.SetCoords(oechem.OEFloatArray(newpos))
oechem.OESetSDData(oe_mol, "Energy FFXML", str(energy))
oechem.OEWriteConstMolecule(ofs, oe_mol)
return
def dock_molecule(molecule, ofs, default_receptor='x0387'):
"""
Dock the specified molecules, writing out to specified file
Parameters
----------
molecule : OEMol
The molecule to dock
ofs : oechem.oemolostream
The filename to stream docked molecules to
default_receptor : str, optional, default='0387'
The default receptor to dock to
"""
# Make a copy of the molecule
molecule = oechem.OEMol(molecule)
import os
# Extract list of corresponding receptor(s)
import oechem
print(f'\n{molecule.GetTitle()}')
if oechem.OEHasSDData(molecule, "fragments"):
fragments = oechem.OEGetSDData(molecule, "fragments").split(',')
print(f'fragments before filter: {fragments}')
fragments = [
fragment for fragment in fragments
if os.path.exists(f'../receptors/Mpro-{fragment}-receptor.oeb.gz')
]
print(f'fragments after filter: {fragments}')
if len(fragments) == 0:
fragments = [default_receptor]
for fragment in fragments:
molecule_to_dock = oechem.OEMol(molecule)
import os
receptor_filename = os.path.join(
f'../receptors/Mpro-{fragment}-receptor.oeb.gz')
oechem.OESetSDData(molecule_to_dock, "fragments", fragment)
# Enumerate reasonable protomers/tautomers
from openeye import oequacpac
protomer = oechem.OEMol()
protomers = [
oechem.OEMol(protomer) for protomer in
oequacpac.OEGetReasonableProtomers(molecule_to_dock)
]
dock_molecules_to_receptor(receptor_filename, protomers, ofs)
def mols_from_file(mol_file):
"""
Parses a standard molecule file into chemper molecules using OpenEye toolkits
Parameters
----------
mol_file: str
relative or full path to molecule containing the molecule file
that is accessible from the current working directory
Returns
-------
mols: list of chemper Mols
list of molecules in the mol2 file as chemper Mols
"""
import os
if not os.path.exists(mol_file):
from chemper.chemper_utils import get_data_path
mol_path = get_data_path(os.path.join('molecules', mol_file))
if not os.path.exists(mol_path):
raise IOError(
"File '%s' not found locally or in chemper/data/molecules." %
mol_file)
else:
mol_file = mol_path
molecules = list()
# make Openeye input file stream
ifs = oechem.oemolistream(mol_file)
oemol = oechem.OECreateOEGraphMol()
while oechem.OEReadMolecule(ifs, oemol):
# if an SD file, the molecule name may be in the SD tags
if oemol.GetTitle() == '':
name = oechem.OEGetSDData(oemol, 'name').strip()
oemol.SetTitle(name)
# Append to list.
molecules.append(Mol(oechem.OEMol(oemol)))
ifs.close()
return molecules
def main(infile):
# open multi-molecule, multi-conformer file
ifs = oechem.oemolistream()
ifs.SetConfTest(oechem.OEAbsCanonicalConfTest())
if not ifs.open(infile):
raise FileNotFoundError(f"Unable to open {infile} for reading")
mols = ifs.GetOEMols()
for i, mol in enumerate(mols):
# perceive stereochemistry for mol
oechem.OEPerceiveChiral(mol)
oechem.OEAssignAromaticFlags(mol, oechem.OEAroModel_MDL)
# assign charges to copy of mol
# note that chg_mol does NOT have conformers
try:
chg_mol = charge_mol(mol)
except RuntimeError:
# perceive stereochem
#find_unspecified_stereochem(mol)
oechem.OE3DToInternalStereo(mol)
# reset perceived and call OE3DToBondStereo, since it may be missed
# by OE3DToInternalStereo if it thinks mol is flat
mol.ResetPerceived()
oechem.OE3DToBondStereo(mol)
try:
chg_mol = charge_mol(mol)
print(f'fixed stereo: {mol.GetTitle()}')
except RuntimeError:
find_unspecified_stereochem(mol)
title = mol.GetTitle()
smilabel = oechem.OEGetSDData(mol, "SMILES QCArchive")
print(' >>> Charge assignment failed due to unspecified '
f'stereochemistry {title} {smilabel}')
continue
def get_fragment_to_parent_atom_mapping(parent_mol, frag_mol):
try:
mapping_data = oechem.OEGetSDData(frag_mol,
FRAGMENT_TO_PARENT_ATOMS_KEY)
idx_map = dict(
map(int, idx_pair.split('_'))
for idx_pair in mapping_data.split('-'))
atom_map = {}
for frag_idx, parent_idx in idx_map.items():
frag_atom = frag_mol.GetAtom(oechem.OEHasAtomIdx(frag_idx - 1))
parent_atom = parent_mol.GetAtom(
oechem.OEHasAtomIdx(parent_idx - 1))
if frag_atom is not None and parent_atom is not None:
atom_map[frag_atom] = parent_atom
return atom_map
except Exception as e:
logging.warning(e)
return {}
def reorder_sd_props(mol: oechem.OEGraphMol):
strain1 = oechem.OEGetSDData(mol, TOTAL_STRAIN_TAG)
data_pairs = [(TOTAL_STRAIN_TAG, strain1)]
for dp in oechem.OEGetSDDataPairs(mol):
if dp.GetTag() == TOTAL_STRAIN_TAG:
pass
else:
data_pairs.append((dp.GetTag(), dp.GetValue()))
oechem.OEClearSDData(mol)
for k, v in data_pairs:
oechem.OESetSDData(mol, k, v)
data_pairs = []
for dp in oechem.OEGetSDDataPairs(mol):
data_pairs.append((dp.GetTag(), dp.GetValue()))
oechem.OEClearSDData(mol)
for k, v in data_pairs:
oechem.OESetSDData(mol, k, v)