def mol_coords_to_numpy_array(mol):
coords = OEFloatArray(mol.GetMaxAtomIdx() * 3)
mol.GetCoords(coords)
arr = np.ctypeslib.as_array(ctypes.cast(int(coords.PtrCast()),
ctypes.POINTER(ctypes.c_float)),
shape=(len(coords), ))
return np.array(arr.reshape((-1, 3)))
def get_biomt(pdb):
"""
"""
statement = text("""
SELECT assembly_serial, assembly_size, is_monomeric,
pdb_chain_id, rotation, translation, is_at_identity
FROM pdb_dev.biomt b
JOIN pdb_dev.biomt_ops o USING(biomt_id)
WHERE pdb = :pdb
ORDER BY 1,3
""")
# fetch data from pisa database
result = engine.execute(statement, pdb=pdb.upper()).fetchall()
biomt = {}
is_monomeric = False
# iterate through assemblies
for (assembly_serial, assembly_size,
is_monomeric), chain_iter in groupby(result, key=itemgetter(0, 1, 2)):
biomt[assembly_serial] = {}
# do not return pisa data for large assemblies / asu will be used instead
if assembly_size > 26:
app.log.warn(
"one of the predicted assemblies contains {} chains - "
"the asymmetric unit will be used instead.".format(
assembly_size))
biomt = {}
break
# the complete ASU is monomeric and has to be split into individual chains
if is_monomeric: break
# iterate through chains
for pdb_chain_id, operation_iter in groupby(chain_iter,
key=itemgetter(3)):
biomt[assembly_serial].update({str(pdb_chain_id): {}})
for operation_serial, operation in enumerate(operation_iter, 1):
rotation, translation, is_at_identity = operation[4:]
details = {
'rotation': OEFloatArray(rotation),
'translation': OEFloatArray(translation),
'is_at_identity': is_at_identity
}
biomt[assembly_serial][pdb_chain_id][
operation_serial] = details
# debug assembly information
try:
app.log.debug("BIOMT contains {0} assembly/assemblies.".format(
max(biomt.keys())))
# PDB entry does not have a REMARK 350
except ValueError:
app.log.debug("NO REMARK 350 found.")
return biomt, is_monomeric
def write_xyz(ofs, mol, coords):
mol.SetCoords(OEFloatArray(coords.reshape(-1)))
oechem.OEWriteMolecule(ofs, mol)
def __getMiscFile(self, ccPath, suppressHydrogens=False, importType="2D"):
"""Fetch a miscellaneous chemical file (ccPath) and build OE molecules
for comparison and depiction.
"""
try:
oem = OeBuildMol(verbose=self.__verbose, log=self.__lfh)
if oem.importFile(ccPath, type=importType):
if self.__verbose:
self.__lfh.write("+OEAlignDepilsct.__getMiscFile()\n")
self.__lfh.write(" Title = %s\n" %
oem.getTitle())
self.__lfh.write(" SMILES = %s\n" %
oem.getCanSMILES())
self.__lfh.write(" SMILES (stereo) = %s\n" %
oem.getIsoSMILES())
self.__lfh.write(" Formula (Hill) = %s\n" %
oem.getFormula())
self.__lfh.write(" InChI key = %s\n" %
oem.getInChIKey())
self.__lfh.write(" InChI = %s\n" %
oem.getInChI())
else:
self.__lfh.write(
"+OEAlignDepict.__getMiscFile() Read failed for %s\n" %
ccPath)
return None, None, None
#
# oem.build2D()
ccId = oem.getTitle()
if suppressHydrogens:
tMol = oem.getGraphMolSuppressH()
else:
tMol = oem.getMol()
molXyzL = []
if importType == "3D":
for ii, atm in enumerate(tMol.GetAtoms()):
xyzL = OEFloatArray(3)
tMol.GetCoords(atm, xyzL)
molXyzL.append(
(ii, atm.GetIdx(), atm.GetAtomicNum(), atm.GetName(),
atm.GetType(), "%.3f" % xyzL[0], "%.3f" % xyzL[1],
"%.3f" % xyzL[2]))
fD = {}
fD = {
"Formula": oem.getFormula(),
"SMILES": oem.getCanSMILES(),
"SMILES_STEREO": oem.getIsoSMILES(),
"InChI": oem.getInChI(),
"InChIKey": oem.getInChIKey(),
"xyz": molXyzL,
}
for ii, atm in enumerate(tMol.GetAtoms()):
xyzL = OEFloatArray(3)
tMol.GetCoords(atm, xyzL)
if self.__verbose:
self.__lfh.write(
"OeAlignDepict.__getMiscFile - atom %d %s %s %s %s %r\n"
% (ii, atm.GetIdx(), atm.GetAtomicNum(), atm.GetName(),
atm.GetType(), xyzL))
return (ccId, tMol, fD)
except: # noqa: E722 pylint: disable=bare-except
traceback.print_exc(file=self.__lfh)
# self.fail()
return None, None, None
def importFile(self, filePath, type='2D'): # pylint: disable=redefined-builtin
""" Contruct a OEGraphMol using the content of the input file. The input
file must have a file extension recognized by the OE toolkit (e.g. .sdf)
"""
ifs = oemolistream()
if not ifs.open(filePath):
return False
#
# self.__oeMol = OEGraphMol()
self.__oeMol = OEMol()
OEReadMolecule(ifs, self.__oeMol)
# OETriposAtomNames(self.__oeMol)
if type == '2D':
# run standard perceptions --
OEFindRingAtomsAndBonds(self.__oeMol)
OEPerceiveChiral(self.__oeMol)
for oeAt in self.__oeMol.GetAtoms():
st = oeAt.GetStringData("StereoInfo")
if st == 'R':
OESetCIPStereo(self.__oeMol, oeAt, OECIPAtomStereo_R)
elif st == 'S':
OESetCIPStereo(self.__oeMol, oeAt, OECIPAtomStereo_S)
for oeBnd in self.__oeMol.GetBonds():
st = oeBnd.GetStringData("StereoInfo")
if st == 'E':
OESetCIPStereo(self.__oeMol, oeBnd, OECIPBondStereo_E)
elif st == 'Z':
OESetCIPStereo(self.__oeMol, oeBnd, OECIPBondStereo_Z)
elif type == '3D':
# run standard perceptions --
#
self.__oeMol.SetDimension(3)
OE3DToInternalStereo(self.__oeMol)
OEFindRingAtomsAndBonds(self.__oeMol)
# Other aromatic models: OEAroModelMDL or OEAroModelDaylight
OEAssignAromaticFlags(self.__oeMol, OEAroModelOpenEye)
self.updateCIPStereoOE()
OEAddExplicitHydrogens(self.__oeMol)
self.__molXyzL = []
aC = {}
for ii, atm in enumerate(self.__oeMol.GetAtoms()):
iAtNum = atm.GetAtomicNum()
if iAtNum in aC:
aC[iAtNum] += 1
else:
aC[iAtNum] = 1
# Less than idea - should have an API
atName = PdbxChemCompConstants._periodicTable[iAtNum - 1] + str(
aC[iAtNum]) # pylint: disable=protected-access
atm.SetName(atName)
#
xyzL = OEFloatArray(3)
self.__oeMol.GetCoords(atm, xyzL)
self.__molXyzL.append(
(ii, atm.GetIdx(), atm.GetAtomicNum(), atm.GetName(),
atm.GetType(), xyzL[0], xyzL[1], xyzL[2]))
return True