def depict_smiles(smiles):
""" OEChem and OEDepict image generation """
# Image to draw on
image = OEImage(400, 400)
# Process SMILES
mol = OEGraphMol()
parsed = OESmilesToMol(mol, str(unquote(smiles)))
if parsed:
# Create image of molecule
newtitle = []
for c in mol.GetTitle():
newtitle.append(choice(ALL_EMOJI))
mol.SetTitle(("".join(newtitle)).encode("UTF-8"))
OEPrepareDepiction(mol)
disp = OE2DMolDisplay(mol)
for adisp in disp.GetAtomDisplays():
adisp.SetLabel(choice(ALL_EMOJI).encode("UTF-8"))
OERenderMolecule(image, disp)
else:
# Create error image
font = OEFont(OEFontFamily_Helvetica, OEFontStyle_Default, 20,
OEAlignment_Center, OERed)
image.DrawText(OE2DPoint(image.GetWidth()/2.0, image.GetHeight()/2.0),
'Your SMILES is not valid', font)
img_content = OEWriteImageToString('svg', image)
return Response(img_content, mimetype='image/svg+xml')
def test_sanitizeSMILES():
"""
Test SMILES sanitization.
"""
smiles_list = [
'CC', 'CCC', '[H][C@]1(NC[C@@H](CC1CO[C@H]2CC[C@@H](CC2)O)N)[H]'
]
sanitized_smiles_list = sanitizeSMILES(smiles_list, mode='drop')
if len(sanitized_smiles_list) != 2:
raise Exception(
"Molecules with undefined stereochemistry are not being properly dropped (size=%d)."
% len(sanitized_smiles_list))
sanitized_smiles_list = sanitizeSMILES(smiles_list, mode='expand')
if len(sanitized_smiles_list) != 4:
raise Exception(
"Molecules with undefined stereochemistry are not being properly expanded (size=%d)."
% len(sanitized_smiles_list))
# Check that all molecules can be round-tripped.
from openeye.oechem import OEGraphMol, OESmilesToMol, OECreateIsoSmiString
for smiles in sanitized_smiles_list:
molecule = OEGraphMol()
OESmilesToMol(molecule, smiles)
isosmiles = OECreateIsoSmiString(molecule)
if (smiles != isosmiles):
raise Exception(
"Molecule '%s' was not properly round-tripped (result was '%s')"
% (smiles, isosmiles))
def test_success(self):
mol = OEMol()
OESmilesToMol(mol, str("c1(c(cccc1)OC(=O)C)C(=O)O"))
self.cube.process(mol, self.cube.intake.name)
self.assertEqual(self.runner.outputs["success"].qsize(), 1)
self.assertEqual(self.runner.outputs["failure"].qsize(), 1)
new_mol = self.runner.outputs["success"].get()
self.assertTrue(OEExactGraphMatch(new_mol, mol))
def sanitizeSMILES(smiles_list, mode='drop', verbose=False):
"""
Sanitize set of SMILES strings by ensuring all are canonical isomeric SMILES.
Duplicates are also removed.
Parameters
----------
smiles_list : iterable of str
The set of SMILES strings to sanitize.
mode : str, optional, default='drop'
When a SMILES string that does not correspond to canonical isomeric SMILES is found, select the action to be performed.
'exception' : raise an `Exception`
'drop' : drop the SMILES string
'expand' : expand all stereocenters into multiple molecules
verbose : bool, optional, default=False
If True, print verbose output.
Returns
-------
sanitized_smiles_list : list of str
Sanitized list of canonical isomeric SMILES strings.
Examples
--------
Sanitize a simple list.
>>> smiles_list = ['CC', 'CCC', '[H][C@]1(NC[C@@H](CC1CO[C@H]2CC[C@@H](CC2)O)N)[H]']
Throw an exception if undefined stereochemistry is present.
>>> sanitized_smiles_list = sanitizeSMILES(smiles_list, mode='exception')
Traceback (most recent call last):
...
Exception: Molecule '[H][C@]1(NC[C@@H](CC1CO[C@H]2CC[C@@H](CC2)O)N)[H]' has undefined stereocenters
Drop molecules iwth undefined stereochemistry.
>>> sanitized_smiles_list = sanitizeSMILES(smiles_list, mode='drop')
>>> len(sanitized_smiles_list)
2
Expand molecules iwth undefined stereochemistry.
>>> sanitized_smiles_list = sanitizeSMILES(smiles_list, mode='expand')
>>> len(sanitized_smiles_list)
4
"""
from openeye import oechem
from openeye.oechem import OEGraphMol, OESmilesToMol, OECreateIsoSmiString
from perses.tests.utils import has_undefined_stereocenters, enumerate_undefined_stereocenters
sanitized_smiles_set = set()
OESMILES_OPTIONS = oechem.OESMILESFlag_DEFAULT | oechem.OESMILESFlag_ISOMERIC | oechem.OESMILESFlag_Hydrogens ## IVY
for smiles in smiles_list:
molecule = OEGraphMol()
OESmilesToMol(molecule, smiles)
oechem.OEAddExplicitHydrogens(molecule)
if verbose:
molecule.SetTitle(smiles)
oechem.OETriposAtomNames(molecule)
if has_undefined_stereocenters(molecule, verbose=verbose):
if mode == 'drop':
if verbose:
print("Dropping '%s' due to undefined stereocenters." % smiles)
continue
elif mode == 'exception':
raise Exception("Molecule '%s' has undefined stereocenters" % smiles)
elif mode == 'expand':
if verbose:
print('Expanding stereochemistry:')
print('original: %s', smiles)
molecules = enumerate_undefined_stereocenters(molecule, verbose=verbose)
for molecule in molecules:
smiles_string = oechem.OECreateSmiString(molecule, OESMILES_OPTIONS) ## IVY
sanitized_smiles_set.add(smiles_string) ## IVY
if verbose: print('expanded: %s', smiles_string)
else:
# Convert to OpenEye's canonical isomeric SMILES.
smiles_string = oechem.OECreateSmiString(molecule, OESMILES_OPTIONS) ## IVY
sanitized_smiles_set.add(smiles_string) ## IVY
sanitized_smiles_list = list(sanitized_smiles_set)
return sanitized_smiles_list
