def _hydrogenize(block, hydro):
mol = Chem.MolFromMolBlock(block, sanitize=False)
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
res = Chem.AddHs(mol, addCoords=True) if hydro else Chem.RemoveHs(
mol, sanitize=False)
return MolToMarvin(Chem.MolToMolBlock(res))
def _similarityMap(ms,
width=500,
height=500,
radius=2,
fingerprint='morgan',
format='png'):
if matplotlib is None:
raise ValueError('matplotlib not useable')
_call(ms, 'UpdatePropertyCache', strict=False)
_apply(ms, ct._sssr)
fn = None
if fingerprint == 'morgan':
fn = lambda x, i: SimilarityMaps.GetMorganFingerprint(
x, i, radius=radius)
elif fingerprint == 'tt':
fn = SimilarityMaps.GetAPFingerprint
elif fingerprint == 'ap':
fn = SimilarityMaps.GetTTFingerprint
SimilarityMaps.GetSimilarityMapForFingerprint(ms[0],
ms[1],
fn,
size=(width, height))
sio = StringIO.StringIO()
pyplot.savefig(sio, format=format, bbox_inches='tight', dpi=100)
return sio.getvalue()
def _autoDetect(structure):
mol = None
if Chem.MolFromSmiles(structure.strip(), sanitize=False):
mol = Chem.MolFromSmiles(structure.strip(), sanitize=False)
elif Chem.MolFromMolBlock(structure, sanitize=False):
return Chem.MolFromMolBlock(structure, sanitize=False)
elif Chem.INCHI_AVAILABLE and Chem.inchi.MolFromInchi(structure.strip(), sanitize=True, removeHs=True):
mol = Chem.inchi.MolFromInchi(structure.strip(), sanitize=True, removeHs=True)
elif hasattr(Chem, 'MolFromSmarts') and Chem.MolFromSmarts(structure):
mol = Chem.MolFromSmarts(structure)
elif hasattr(Chem, 'MolFromMol2Block') and Chem.MolFromMol2Block(structure):
mol = Chem.MolFromMol2Block(structure)
elif hasattr(Chem, 'MolFromPDBBlock') and Chem.MolFromPDBBlock(structure):
mol = Chem.MolFromPDBBlock(structure)
elif hasattr(Chem, 'MolFromTPLBlock') and Chem.MolFromTPLBlock(structure):
mol = Chem.MolFromTPLBlock(structure)
else:
try:
mol = Chem.MolFromMolBlock(MarvinToMol(structure))
except:
pass
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
return mol
def _mols2imageStream(mols, f, format, size, legend, highlightAtomLists=None, kekulize=True):
labels = [x for x in islice(cycle(legend), len(mols))] if isinstance(legend, (list, tuple)) else \
[x for x in islice(cycle([legend]), len(mols))]
legends = [x.GetProp("_Name") if (x.HasProp("_Name") and x.GetProp("_Name")) else labels[idx]
for idx, x in enumerate(mols)]
molsPerRow = min(len(mols), 4)
nRows = len(mols) // molsPerRow
if len(mols) % molsPerRow:
nRows += 1
_call(mols, 'UpdatePropertyCache', strict=False)
if kekulize:
_apply(mols, _kekulize)
highlightBondLists = []
if highlightAtomLists:
for mol, highlightAtomList in zip(mols, highlightAtomLists):
highlightBondList = []
for bnd in mol.GetBonds():
if bnd.GetBeginAtomIdx() in highlightAtomList and bnd.GetEndAtomIdx() in highlightAtomList:
highlightBondList.append(bnd.GetIdx())
highlightBondLists.append(highlightBondList)
if not highlightBondLists:
highlightBondLists = None
panelx = size
panely = size
canvasx = panelx * molsPerRow
canvasy = panely * nRows
drawer = rdMolDraw2D.MolDraw2DCairo(canvasx, canvasy, panelx, panely)
drawer.DrawMolecules(mols, highlightAtoms=highlightAtomLists, highlightBonds=highlightBondLists, legends=legends)
drawer.FinishDrawing()
f.write(drawer.GetDrawingText())
def _isBondInRing(data,
index,
size,
sanitize=True,
removeHs=True,
strictParsing=True):
mols = _parseMolData(data,
sanitize=sanitize,
removeHs=removeHs,
strictParsing=strictParsing)
return _call(_call(mols, 'GetRingInfo'), 'IsBondInRingOfSize', index, size)
def _clean(mrv, dim=2):
block = MarvinToMol(mrv)
mol = Chem.MolFromMolBlock(block, sanitize=False)
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
if not mol:
print "No mol for block:\n %s" % block
return mrv
AllChem.Compute2DCoords(mol, bondLength=0.8)
if dim == 3:
mol = _2D23D(mol, True)
mol = Chem.RemoveHs(mol)
return MolToMarvin(Chem.MolToMolBlock(mol))
def _clean(mrv, dim=2):
block = MarvinToMol(mrv)
mol = Chem.MolFromMolBlock(block, sanitize=False)
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
if not mol:
print "No mol for block:\n %s" % block
return mrv
AllChem.Compute2DCoords(mol, bondLength=0.8)
if dim == 3:
mol = _2D23D(mol, True)
mol = Chem.RemoveHs(mol)
return MolToMarvin(Chem.MolToMolBlock(mol))
def _parseSMILESData(data, computeCoords=False, delimiter=' ', smilesColumn=0, nameColumn=1, titleLine=True,
sanitize=True):
fd, fpath = tempfile.mkstemp(text=True)
os.write(fd, data)
os.close(fd)
suppl = SmilesMolSupplier(fpath, delimiter=delimiter, smilesColumn=smilesColumn, nameColumn=nameColumn,
titleLine=titleLine, sanitize=sanitize)
mols = [x for x in suppl if x]
# if not mols:
# mols = [MolFromSmiles(data, sanitize=sanitize)]
if computeCoords:
_apply(mols, _computeCoords, True)
_call(mols, 'SetProp', '_Name', '')
os.remove(fpath)
return mols
def _numBondRings(data, sanitize=True, removeHs=True, strictParsing=True):
mols = _parseMolData(data,
sanitize=sanitize,
removeHs=removeHs,
strictParsing=strictParsing)
ring_infos = _call(mols, 'GetRingInfo')
return [[ring_info.NumBondRings(bond.GetIdx()) for bond in mol.GetBonds()]
for (mol, ring_info) in zip(mols, ring_infos)]
def _molExport(structure, **kwargs):
input_f = kwargs.get('input', None)
output_f = kwargs.get('output', None)
if not input_f:
mol = _autoDetect(str(structure))
elif input_f == 'mrv':
mol = Chem.MolFromMolBlock(MarvinToMol(structure), sanitize=False)
elif input_f == 'smiles':
mol = Chem.MolFromSmiles(str(structure), sanitize=False)
elif input_f == 'inchi':
mol = Chem.MolFromInchi(str(structure))
else:
mol = Chem.MolFromMolBlock(structure, sanitize=False)
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
if not mol.GetNumConformers() or mol.GetConformer().Is3D():
AllChem.Compute2DCoords(mol, bondLength=0.8)
if output_f == 'smiles':
out_structure = Chem.MolToSmiles(mol)
elif output_f == 'inchi':
out_structure = Chem.MolToInchi(mol)
elif output_f == 'inchikey':
out_structure = Chem.InchiToInchiKey(Chem.MolToInchi(mol))
elif output_f == 'mol':
out_structure = Chem.MolToMolBlock(mol)
elif output_f == 'sdf':
out_structure = Chem.MolToMolBlock(mol) + '\n$$$$\n'
elif output_f == 'mrv':
out_structure = MolToMarvin(Chem.MolToMolBlock(mol))
return {
"structure": out_structure,
"format": output_f,
"contentUrl": "",
"contentBaseUrl": ""
}
def _mols2svg(mols, size, legend, kekulize=True, wedgeBonds=True, fitImage=True, atomMapNumber=False,
computeCoords=False, highlightAtomLists=None):
if not mols:
return ''
_call(mols, 'UpdatePropertyCache', strict=False)
if computeCoords:
_apply(mols, _computeCoords, True)
if atomMapNumber:
_apply(mols, _atomMapNumber)
labels = [x for x in islice(cycle(legend), len(mols))] if isinstance(legend, (list, tuple)) else \
[x for x in islice(cycle([legend]), len(mols))]
legends = [x.GetProp("_Name") if (x.HasProp("_Name") and x.GetProp("_Name")) else labels[idx]
for idx, x in enumerate(mols)]
molsPerRow = min(len(mols), 4)
nRows = len(mols) // molsPerRow
if len(mols) % molsPerRow:
nRows += 1
if kekulize:
_apply(mols, _kekulize)
highlightBondLists = []
if highlightAtomLists:
for mol, highlightAtomList in zip(mols, highlightAtomLists):
highlightBondList = []
for bnd in mol.GetBonds():
if bnd.GetBeginAtomIdx() in highlightAtomList and bnd.GetEndAtomIdx() in highlightAtomList:
highlightBondList.append(bnd.GetIdx())
highlightBondLists.append(highlightBondList)
if not highlightBondLists:
highlightBondLists = None
panelx = size
panely = size
canvasx = panelx * molsPerRow
canvasy = panely * nRows
drawer = rdMolDraw2D.MolDraw2DSVG(canvasx, canvasy, panelx, panely)
drawer.DrawMolecules(mols, highlightAtoms=highlightAtomLists, highlightBonds=highlightBondLists, legends=legends)
drawer.FinishDrawing()
return drawer.GetDrawingText()
def _similarityMap(ms, width=500, height=500, radius=2, fingerprint='morgan', format='png'):
if matplotlib is None:
raise ValueError('matplotlib not useable')
_call(ms, 'UpdatePropertyCache', strict=False)
_apply(ms, ct._sssr)
fn = None
if fingerprint == 'morgan':
fn = lambda x, i: SimilarityMaps.GetMorganFingerprint(x, i, radius=radius)
elif fingerprint == 'tt':
fn = SimilarityMaps.GetAPFingerprint
elif fingerprint == 'ap':
fn = SimilarityMaps.GetTTFingerprint
SimilarityMaps.GetSimilarityMapForFingerprint(ms[0], ms[1], fn, size=(width, height))
sio = StringIO.StringIO()
pyplot.savefig(sio, format=format, bbox_inches='tight', dpi=100)
return sio.getvalue()
def _molExport(structure, **kwargs):
input_f = kwargs.get('input', None)
output_f = kwargs.get('output', None)
if not input_f:
mol = _autoDetect(str(structure))
elif input_f == 'mrv':
mol = Chem.MolFromMolBlock(MarvinToMol(structure), sanitize=False)
elif input_f == 'smiles':
mol = Chem.MolFromSmiles(str(structure), sanitize=False)
elif input_f == 'inchi':
mol = Chem.MolFromInchi(str(structure))
else:
mol = Chem.MolFromMolBlock(structure, sanitize=False)
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
if not mol.GetNumConformers() or mol.GetConformer().Is3D():
AllChem.Compute2DCoords(mol, bondLength=0.8)
if output_f == 'smiles':
out_structure = Chem.MolToSmiles(mol)
elif output_f == 'inchi':
out_structure = Chem.MolToInchi(mol)
elif output_f == 'inchikey':
out_structure = Chem.InchiToInchiKey(Chem.MolToInchi(mol))
elif output_f == 'mol':
out_structure = Chem.MolToMolBlock(mol)
elif output_f == 'sdf':
out_structure = Chem.MolToMolBlock(mol) + '\n$$$$\n'
elif output_f == 'mrv':
out_structure = MolToMarvin(Chem.MolToMolBlock(mol))
return {"structure": out_structure, "format": output_f, "contentUrl": "", "contentBaseUrl": ""}
def _autoDetect(structure):
mol = None
if Chem.MolFromSmiles(structure.strip(), sanitize=False):
mol = Chem.MolFromSmiles(structure.strip(), sanitize=False)
elif Chem.MolFromMolBlock(structure, sanitize=False):
return Chem.MolFromMolBlock(structure, sanitize=False)
elif Chem.INCHI_AVAILABLE and Chem.inchi.MolFromInchi(
structure.strip(), sanitize=True, removeHs=True):
mol = Chem.inchi.MolFromInchi(structure.strip(),
sanitize=True,
removeHs=True)
elif hasattr(Chem, 'MolFromSmarts') and Chem.MolFromSmarts(structure):
mol = Chem.MolFromSmarts(structure)
elif hasattr(Chem,
'MolFromMol2Block') and Chem.MolFromMol2Block(structure):
mol = Chem.MolFromMol2Block(structure)
elif hasattr(Chem, 'MolFromPDBBlock') and Chem.MolFromPDBBlock(structure):
mol = Chem.MolFromPDBBlock(structure)
elif hasattr(Chem, 'MolFromTPLBlock') and Chem.MolFromTPLBlock(structure):
mol = Chem.MolFromTPLBlock(structure)
else:
try:
mol = Chem.MolFromMolBlock(MarvinToMol(structure))
except:
pass
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
return mol
def _parseSMILESData(data,
computeCoords=False,
delimiter=' ',
smilesColumn=0,
nameColumn=1,
titleLine=True,
sanitize=True):
fd, fpath = tempfile.mkstemp(text=True)
os.write(fd, data)
os.close(fd)
suppl = SmilesMolSupplier(fpath,
delimiter=delimiter,
smilesColumn=smilesColumn,
nameColumn=nameColumn,
titleLine=titleLine,
sanitize=sanitize)
mols = [x for x in suppl if x]
# if not mols:
# mols = [MolFromSmiles(data, sanitize=sanitize)]
if computeCoords:
_apply(mols, _computeCoords, True)
_call(mols, 'SetProp', '_Name', '')
os.remove(fpath)
return mols
def _align(mols, pattern, force=False):
if all(mol.HasSubstructMatch(pattern) for mol in mols):
_apply(mols, GenerateDepictionMatching2DStructure, pattern)
return mols
if not force:
return
pattern = _adjustQuery(pattern)
matches = _call(mols, 'HasSubstructMatch', pattern)
if all(matches):
_apply(mols, GenerateDepictionMatching2DStructure, pattern)
return mols
if any(matches):
_apply(mols, GenerateDepictionMatching2DStructure, pattern, acceptFailure=True)
return compress(mols, matches)
def _align(mols, pattern, force=False):
if all(mol.HasSubstructMatch(pattern) for mol in mols):
_apply(mols, GenerateDepictionMatching2DStructure, pattern)
return mols
if not force:
return
pattern = _adjustQuery(pattern)
matches = _call(mols, 'HasSubstructMatch', pattern)
if all(matches):
_apply(mols, GenerateDepictionMatching2DStructure, pattern)
return mols
if any(matches):
_apply(mols,
GenerateDepictionMatching2DStructure,
pattern,
acceptFailure=True)
return compress(mols, matches)
def _getMatches(mols, smarts, force=False):
_call(mols, 'UpdatePropertyCache', strict=False)
return _apply(mols, _getSubstructMatch, _molFromSmarts(smarts), force)
def _numRings(data, sanitize=True, removeHs=True, strictParsing=True):
mols = _parseMolData(data, sanitize=sanitize, removeHs=removeHs, strictParsing=strictParsing)
return _call(_call(mols, 'GetRingInfo'), 'NumRings')
def _getNumBonds(data, sanitize=True, removeHs=True, strictParsing=True):
return _call(_parseMolData(data, sanitize=sanitize, removeHs=removeHs, strictParsing=strictParsing), "GetNumBonds")
def _hydrogenize(block, hydro):
mol = Chem.MolFromMolBlock(block, sanitize=False)
_call([mol], 'UpdatePropertyCache', strict=False)
_apply([mol], ct._sssr)
res = Chem.AddHs(mol, addCoords=True) if hydro else Chem.RemoveHs(mol, sanitize=False)
return MolToMarvin(Chem.MolToMolBlock(res))
def _getNumAtoms(data):
return _call(_parseMolData(data), "GetNumAtoms")
def _getMatches(mols, smarts, force=False):
_call(mols, 'UpdatePropertyCache', strict=False)
return _apply(mols, _getSubstructMatch, _molFromSmarts(smarts), force)
def _getNumAtoms(data):
return _call(_parseMolData(data), "GetNumAtoms")
def _numBondRings(data, sanitize=True, removeHs=True, strictParsing=True):
mols = _parseMolData(data, sanitize=sanitize, removeHs=removeHs, strictParsing=strictParsing)
ring_infos = _call(mols, 'GetRingInfo')
return [[ring_info.NumBondRings(bond.GetIdx()) for bond in mol.GetBonds()] for (mol, ring_info) in zip(mols, ring_infos)]
def _getNumBonds(data, sanitize=True, removeHs=True, strictParsing=True):
return _call(
_parseMolData(data,
sanitize=sanitize,
removeHs=removeHs,
strictParsing=strictParsing), "GetNumBonds")
def _numRings(data, sanitize=True, removeHs=True, strictParsing=True):
mols = _parseMolData(data,
sanitize=sanitize,
removeHs=removeHs,
strictParsing=strictParsing)
return _call(_call(mols, 'GetRingInfo'), 'NumRings')
def _isBondInRing(data, index, size, sanitize=True, removeHs=True, strictParsing=True):
mols = _parseMolData(data, sanitize=sanitize, removeHs=removeHs, strictParsing=strictParsing)
return _call(_call(mols, 'GetRingInfo'), 'IsBondInRingOfSize', index, size)
def _mols2svg(mols,
size,
legend,
kekulize=True,
wedgeBonds=True,
fitImage=True,
atomMapNumber=False,
computeCoords=False,
highlightAtomLists=None):
if not mols:
return ''
_call(mols, 'UpdatePropertyCache', strict=False)
if computeCoords:
_apply(mols, _computeCoords, True)
if atomMapNumber:
_apply(mols, _atomMapNumber)
labels = [x for x in islice(cycle(legend), len(mols))] if isinstance(legend, (list, tuple)) else \
[x for x in islice(cycle([legend]), len(mols))]
legends = [
x.GetProp("_Name") if
(x.HasProp("_Name") and x.GetProp("_Name")) else labels[idx]
for idx, x in enumerate(mols)
]
molsPerRow = min(len(mols), 4)
nRows = len(mols) // molsPerRow
if len(mols) % molsPerRow:
nRows += 1
if NEW_RENDER_ENGINE:
if kekulize:
_apply(mols, _kekulize)
highlightBondLists = []
if highlightAtomLists:
for mol, highlightAtomList in zip(mols, highlightAtomLists):
highlightBondList = []
for bnd in mol.GetBonds():
if bnd.GetBeginAtomIdx(
) in highlightAtomList and bnd.GetEndAtomIdx(
) in highlightAtomList:
highlightBondList.append(bnd.GetIdx())
highlightBondLists.append(highlightBondList)
if not highlightBondLists:
highlightBondLists = None
panelx = size
panely = size
canvasx = panelx * molsPerRow
canvasy = panely * nRows
drawer = rdMolDraw2D.MolDraw2DSVG(canvasx, canvasy, panelx, panely)
drawer.DrawMolecules(mols,
highlightAtoms=highlightAtomLists,
highlightBonds=highlightBondLists,
legends=legends)
drawer.FinishDrawing()
return drawer.GetDrawingText()
totalWidth = molsPerRow * size
totalHeight = molsPerRow * size
if cffi and cairocffi.version <= (1, 10, 0):
imageData = io.BytesIO()
else:
imageData = StringIO.StringIO()
surf = cairo.SVGSurface(imageData, totalWidth, totalHeight)
ctx = cairo.Context(surf)
for i, mol in enumerate(mols):
highlight = []
if highlightAtomLists:
highlight = highlightAtomLists[i] or []
tx = size * (i % molsPerRow)
ty = size * (i // molsPerRow)
ctx.translate(tx, ty)
canv = cairoCanvas.Canvas(ctx=ctx, size=(size, size), imageType='svg')
draw.MolToImage(mol,
size=(size, size),
legend=legends[i],
canvas=canv,
kekulize=kekulize,
wedgeBonds=wedgeBonds,
fitImage=fitImage,
highlightAtoms=highlight)
canv.flush()
ctx.translate(-tx, -ty)
surf.finish()
return imageData.getvalue()