def append_matrix_with_rmsd_to_previous_element(
matrix, iterator, fieldname_rmsd="RMSD_to_previous"):
import numpy
ligands = []
foobar = None
rmsd = 0
for pos in iterator:
filename = matrix[pos]["filename"]
ligands.append(openbabel.OBMol())
obconversion.SetInFormat(filename[filename.rfind(".") + 1:])
if obconversion.ReadFile(ligands[-1], filename):
log.info('Successfully read ligand ' + str(filename))
else:
log.warning('Failed to read ligand ' + str(filename))
if foobar == None:
foobar = 1 # skip first
continue
else:
#print "hi"
foobar = openbabel.OBMolAtomIter(ligands[-2])
for atom in openbabel.OBMolAtomIter(ligands[-1]):
V = foobar.__next__().GetVector()
W = atom.GetVector()
rmsd += (V.distSq(W))
rmsd /= ligands[-1].NumAtoms()
rmsd = numpy.sqrt(rmsd)
matrix[pos].update({fieldname_rmsd: rmsd})
return
def CheckBondConnectivity(poltype,mol,optmol):
atomitermol=openbabel.OBMolAtomIter(mol)
atomiteroptmol=openbabel.OBMolAtomIter(optmol)
for atm in atomitermol:
atmidxmol=atm.GetIdx()
atmoptmol=optmol.GetAtom(atmidxmol)
atmidxoptmol=atmoptmol.GetIdx()
atmneighbidxlist=[]
iteratomatommol = openbabel.OBAtomAtomIter(atm)
for newatm in iteratomatommol:
atmneighbidxlist.append(newatm.GetIdx())
atmneighbidxlistoptmol=[]
iteratomatomoptmol = openbabel.OBAtomAtomIter(atmoptmol)
for newatm in iteratomatomoptmol:
atmneighbidxlistoptmol.append(newatm.GetIdx())
if set(atmneighbidxlist)!=set(atmneighbidxlistoptmol):
if len(set(atmneighbidxlist))>len(set(atmneighbidxlistoptmol)):
diff=set(atmneighbidxlist)-set(atmneighbidxlistoptmol)
idxset='mol'
else:
diff=set(atmneighbidxlistoptmol)-set(atmneighbidxlist)
idxset='optmol'
RaiseConnectivityError(poltype,diff,idxset)
def ApplyKabsch(molFit, COMFit, COMRef, rot):
crd=[]
for at in openbabel.OBMolAtomIter(molFit):
crd.append([at.GetX()-COMFit[0], at.GetY()-COMFit[1], at.GetZ()-COMFit[2]])
crd=np.array(crd)
for i in xrange(len(crd)):
crd[i]=np.dot(rot,crd[i])
i = 0
for at in openbabel.OBMolAtomIter(molFit):
at.SetVector(crd[i][0]+COMRef[0], crd[i][1]+COMRef[1], crd[i][2]+COMRef[2])
i+=1
return True
def mol2amol(mol):
mol.Kekulize()
atoms = list()
for a in openbabel.OBMolAtomIter(mol):
atom = list()
atom.append(a.GetHyb())
atom.append(a.GetAtomicNum())
atom.append(a.GetFormalCharge())
atom.append(a.GetIsotope())
if a.IsClockwise():
stereo = 1
elif a.IsAntiClockwise():
stereo = 2
elif a.IsChiral():
stereo = 3
else:
stereo = 0
atom.append(stereo)
atom.append(a.GetSpinMultiplicity())
if a.IsAromatic():
aromatic = 1
else:
aromatic = 0
atom.append(aromatic)
atoms.append(atom)
bonds = list()
for b in openbabel.OBMolBondIter(mol):
bond = list()
bond.append(b.GetBeginAtomIdx())
bond.append(b.GetEndAtomIdx())
bond.append(b.GetBondOrder())
if b.IsWedge():
stereo = 1
elif b.IsHash():
stereo = 6
else:
stereo = 0
bond.append(stereo)
if b.IsAromatic():
aromatic = 1
else:
aromatic = 0
bond.append(aromatic)
if b.IsUp():
updown = 1
print
updown
elif b.IsDown():
updown = 2
print
updown
else:
updown = 0
bond.append(updown)
bonds.append(bond)
return pickle.dumps([atoms, bonds], 2)
def get_atnums(self):
if self._atnums is None:
self._atnums = numpy.zeros((self.natoms, ))
for at in openbabel.OBMolAtomIter(self.obmol):
i = at.GetIdx() - 1
self._atnums[i] = at.GetAtomicNum()
return self._atnums
def get_atmasses(self):
if self._atmasses is None:
self._atmasses = numpy.zeros((self.natoms, ))
for at in openbabel.OBMolAtomIter(self.obmol):
i = at.GetIdx() - 1
self._atmasses[i] = at.GetExactMass()
return self._atmasses
def attype_groups_2(self):
groupnames = ['NH', 'CO', 'CA', 'HA', 'CHB']
groups = []
for k in groupnames:
groups.append([])
for at in openbabel.OBMolAtomIter(self.obmol):
attype = at.GetResidue().GetAtomID(at)
attype = attype.strip()
if attype in ['1HB', '2HB', '3HB', 'HB1', 'HB2', 'HB3']:
attype = 'HB'
if attype in ['1H', '2H', 'H 1', 'H 2', 'H1', 'H2']:
attype = 'HXT'
if attype in ['N', 'H', 'HXT']:
attype = 'NH'
elif attype in ['C', 'O', 'OXT']:
attype = 'CO'
elif attype in ['CB', 'HB']:
attype = 'CHB'
ind = groupnames.index(attype)
groups[ind].append(at.GetIdx() - 1)
return groups, groupnames
def ComputeBondResidual(self, db):
for f in self.frags:
# constructing submolecule from the fragment
efpmol = pyEFP.EFP2mol("./"+db.dirname+"/"+f.ename+"/"+f.fname+".efp")
# fit atoms from the core lists
rot, COMRef, COMFit = pyEFP.SolveKabsch(efpmol, f.subMol, f.efpcore, f.molcore)
# deleting extra atoms
complete = 0
while (complete == 0):
complete = 1
for at in openbabel.OBMolAtomIter(efpmol):
extra = 0
if at.GetIndex()+1 not in f.efptotal:
extra = 1
for at1 in openbabel.OBAtomAtomIter(at):
if (at1.GetIndex()+1 in f.efptotal) and (at.GetAtomicNum() == 1):
extra = 0
break
if (extra == 1):
efpmol.DeleteAtom(at)
complete = 0
efpmol.AddHydrogens()
pyEFP.ApplyKabsch(efpmol, COMFit, COMRef, rot)
for i in xrange(len(f.links)):
dx = efpmol.GetAtom(f.links[i][0]).x() - self.frags[f.links[i][2]].subMol.GetAtom(f.links[i][4]).x()
dy = efpmol.GetAtom(f.links[i][0]).y() - self.frags[f.links[i][2]].subMol.GetAtom(f.links[i][4]).y()
dz = efpmol.GetAtom(f.links[i][0]).z() - self.frags[f.links[i][2]].subMol.GetAtom(f.links[i][4]).z()
f.links[i].append(np.sqrt(dx*dx + dy*dy + dz*dz))
def get_connectivity(self):
'''
Retrieve the connectivity matrix of the molecule.
Returns:
numpy.ndarray: (n_atoms x n_atoms) array containing the pairwise bond orders
between atoms (0 for no bond).
'''
if self._connectivity is None:
# get connectivity matrix
connectivity = np.zeros((self.n_atoms, len(self.numbers)))
for atom in ob.OBMolAtomIter(self.get_obmol()):
index = atom.GetIdx() - 1
# loop over all neighbors of atom
for neighbor in ob.OBAtomAtomIter(atom):
idx = neighbor.GetIdx() - 1
bond_order = neighbor.GetBond(atom).GetBO()
#print(f'{index}-{idx}: {bond_order}')
# do not count bonds between two hydrogen atoms
if (self.numbers[index] == 1 and self.numbers[idx] == 1
and bond_order > 0):
bond_order = 0
connectivity[index, idx] = bond_order
self._connectivity = connectivity
return self._connectivity
def enumerateChiral(mol):
""" enumerate tetrahedral stereocenters
in the molecule
"""
if not mol.IsChiral():
yield mol # we know the molecule is not chiral
else:
chiral = []
chiralSet = []
facade = ob.OBStereoFacade(mol)
for a in ob.OBMolAtomIter(mol):
idx = a.GetIdx()
if a.IsChiral():
chiral.append(idx)
if len(chiral) == 0:
yield mol # no chiral centers
else:
atomList = chiral
step = len(atomList)
for winding in itertools.product([1,2], repeat=step):
for idx, aIdx in enumerate(atomList):
ts = facade.GetTetrahedralStereo(aIdx-1)
config = ts.GetConfig()
config.winding = winding[idx]
config.specified = True
ts.SetConfig(config)
yield mol
def to_graph(self, molw: pybel.Molecule) -> nx.Graph:
mol = molw.OBMol
if self._add_hydrogens:
mol.AddHydrogens()
else:
mol.DeleteHydrogens()
g = nx.Graph()
for atom in openbabel.OBMolAtomIter(mol):
node_features = {
'Symbol': self._element_tabel.GetSymbol(atom.GetAtomicNum())
}
g.add_node(atom.GetIdx(), **node_features)
self._node_types.add(node_features['Symbol'])
for b in openbabel.OBMolBondIter(mol):
idx_a = b.GetBeginAtomIdx()
idx_b = b.GetEndAtomIdx()
edge_features = {'BondType': b.GetBondOrder()}
g.add_edge(idx_a, idx_b, **edge_features)
self._edge_types.add(edge_features['BondType'])
# canonical SMILES format
text = molw.write(format='can')
g.graph['smiles'] = text.split('\t', 1)[0]
g.graph['name'] = molw.title.split('\t', 1)[0]
return g
def make_conv_input(mol,A,B,gshape=64):
"""
Takes an openbabel molecule input, mol, and the atom indices for the two atoms for which you are interested in the coupling
returns a flattend numpy array of the electrostatic potential with the X,Y,Z coordinates of A and B appended
"""
epsilon=0.00000001
grid=np.linspace(-6.,6.,gshape)
XXX,YYY,ZZZ=np.meshgrid(grid,grid,grid)
V=np.zeros((gshape,gshape,gshape,2))
atom_iter=ob.OBMolAtomIter(mol)
for atom in atom_iter:
atomic_num=atom.GetAtomicNum()
x=atom.GetX()
y=atom.GetY()
z=atom.GetZ()
RRR=np.sqrt((XXX-x)**2+(YYY-y)**2+(ZZZ-z)**2)
# V[:,:,:,0]=V[:,:,:,0] - atomic_num/(RRR+epsilon)
V[:,:,:,0]=V[:,:,:,0]-atomic_num*np.exp(-0.5*RRR*RRR)
atom=mol.GetAtom(A)
x=atom.GetX()
y=atom.GetY()
z=atom.GetZ()
RRR2=(XXX-x)**2+(YYY-y)**2+(ZZZ-z)**2
V[:,:,:,1]=V[:,:,:,1] +np.exp(-0.5*(RRR2))
atom=mol.GetAtom(B)
x=atom.GetX()
y=atom.GetY()
z=atom.GetZ()
RRR2=(XXX-x)**2+(YYY-y)**2+(ZZZ-z)**2
V[:,:,:,1]=V[:,:,:,1] +np.exp(-0.5*(RRR2))
return V
def getRingAtomsMulti(molfilePath):
for m in pybel.readfile("mol", molfilePath):
#print m.OBMol.GetFormula()
outString = ""
rings = m.OBMol.GetSSSR()
numR = 0
for r in rings:
numR = numR + 1
outString += "Ring count " + str(numR) + "\n"
for r in rings:
outString += "Ring size " + str(r.Size()) + "\n"
path = r._path
for p in path:
outString += str(p - 1) + "\n"
outString += "Atom count " + str(m.OBMol.NumAtoms()) + "\n"
outString += "Index HowManyRings RingSize Hybridization Hydro_count Aromaticity AntiClockwise_chiral" + "\n"
i = 0
for a in openbabel.OBMolAtomIter(m.OBMol):
outString += str(i) + " " + str(a.MemberOfRingCount()) + " " + str(
a.MemberOfRingSize()) + " " + str(a.GetHyb()) + " " + str(
a.ImplicitHydrogenCount()) + " "
if a.IsAromatic():
outString += "1" + " "
else:
outString += "0" + " "
if a.IsClockwise():
outString += "1" + "\n"
else:
outString += "0" + "\n"
i = i + 1
return outString
def assign_atom_types(selection='all'):
"""
TODO document me!
read http://openbabel.org/dev-api/classOpenBabel_1_1OBAtom.shtml#ae09ed28481ac044dab3f31c8605b44a9
for available functions provided by openbabel to extract atom properties. There is a GetType() function
but I found that function very limited.
"""
atom_types = []
pdb_string = cmd.get_pdbstr(selection)
mol = ob.OBMol()
obconversion = ob.OBConversion()
obconversion.SetInAndOutFormats('pdb', 'pdb')
obconversion.ReadString(mol, pdb_string)
rings = mol.GetSSSR()
for at in ob.OBMolAtomIter(mol):
ring_member = [ring.IsMember(at) for ring in rings]
neighbors = [
neighbor.GetAtomicNum() for neighbor in ob.OBAtomAtomIter(at)
]
atom_types.append(
(at.GetIndex(), at.GetAtomicNum(), at.GetHvyValence(),
any(ring_member), at.IsAromatic(), at.MemberOfRingCount(),
(neighbors)))
return atom_types
def CalculateSymmetry(poltype,pmol, frag_atoms, symmetry_classes):
"""
Intent: Uses and builds on openbabel's 'GetGIVector' method which,
"Calculates a set of graph invariant indexes using the graph theoretical distance,
number of connected heavy atoms, aromatic boolean,
ring boolean, atomic number, and summation of bond orders connected to the atom", to
find initial symmetry classes.
Input:
pmol: OBMol object
frag_atoms: OBBitVec object containing information about the largest fragment in 'pmol'
symmetry_classes: the symmetry_classes array which will be filled in
Output:
nclasses: # of classes
symmetry_classes: array is filled in
Referenced By: gen_canonicallabels
Description:
1. vectorUnsignedInt object is created, 'vgi'
2. It is filled in with the call to GetGIVector
3. The 'symmetry_classes' array is initially filled in to match with 'vgi'
4. These initial invariant classes do not suit our needs perfectly,
so the ExtendInvariants method is called to find the more
refined classes that we need
"""
vgi = openbabel.vectorUnsignedInt()
natoms = pmol.NumAtoms()
nfragatoms = frag_atoms.CountBits()
pmol.GetGIVector(vgi)
iteratom = openbabel.OBMolAtomIter(pmol)
for atom in iteratom:
idx = atom.GetIdx()
if(frag_atoms.BitIsOn(idx)):
symmetry_classes.append([atom, vgi[idx-1]])
nclasses = ExtendInvariants(poltype,pmol, symmetry_classes,frag_atoms,nfragatoms,natoms)
return nclasses
def check_charges(
self,
negcharges=False): # negcharges True means that they are allowed
atoms = list(openbabel.OBMolAtomIter(self.obmol))
for atom in self.G.nodes:
a = atom - 1 # G indices are OBMol indices+1
if self.mapping: # means OBMol was loaded from G at some point
a = self.mapping[atom] - 1 # G indices are OBMol indices+1
if self.G.nodes[atom]['Charge'] > 0: # iterate thru charged atoms
if atoms[a].ExplicitHydrogenCount(
): # check if OBMol bonded a hydrogen to the atom to compensate for false charge
self.G.nodes[atom]['Charge'] = 0 # reset charge in self.G
self.obmol.BeginModify(
) # need this to make changes to existing OBMol object
atoms[a].SetFormalCharge(
0
) # delete charge (yes this will cause problems for >+1, but best method I could come up with)
self.obmol.EndModify(
) # need this to save changes to existing OBMol object
else: # set charge to 0, check valences, re-adjust
if atoms[a].ImplicitHydrogenCount() > 0:
self.G.nodes[atom]['Charge'] = 0
self.obmol.BeginModify()
atoms[a].SetFormalCharge(0)
self.obmol.EndModify()
if self.G.nodes[atom][
'Charge'] < 0 and negcharges is False: # means we will remove this neg charge
self.G.nodes[atom]['Charge'] = 0 # set to 0
self.remove_hydrogens() # remove hydrogens that aren't needed anymore
self.update() # reset OBMol, SMILES, and pybelMol
return
def process_types(poltype, mol):
"""
Intent: Set up scalelist array for scaling certain multipole values
For alchol, the quadrupole on O and H should be mannually scaled by 0.6. This only applies to OH that connect to sp3 Carbon. Similarly for NH in amine (that connects to a sp3 C), scale the Q by 0.75 or 75%. See JCC 2011 32(5):967-77.
"""
scalelist = {}
multipole_scale_dict = {}
for atm in openbabel.OBMolAtomIter(mol):
if symm.get_class_number(poltype, atm.GetIdx()) not in scalelist:
scalelist[symm.get_class_number(poltype, atm.GetIdx())] = []
scalelist[symm.get_class_number(poltype,
atm.GetIdx())].append(None)
scalelist[symm.get_class_number(poltype,
atm.GetIdx())].append(None)
scalelist[symm.get_class_number(poltype,
atm.GetIdx())].append(None)
multipole_scale_dict = {}
#multipole_scale_dict['[OH][CX4]'] = [2, 0.6]
#multipole_scale_dict['[NH2][CX4]'] = [2, 0.75]
for (sckey, scval) in multipole_scale_dict.items():
sp = openbabel.OBSmartsPattern()
openbabel.OBSmartsPattern.Init(sp, sckey)
match = sp.Match(mol)
for ia in sp.GetUMapList():
scalelist[symm.get_class_number(poltype,
ia[0])][scval[0]] = scval[1]
return scalelist
def FitCoords(self, db):
Fitmol = openbabel.OBMol()
for f in self.frags:
# constructing submolecule from the fragment
efpmol = pyEFP.EFP2mol("./"+db.dirname+"/"+f.ename+"/"+f.fname+".efp")
# fit atoms from the core lists
sys.stdout.flush()
rot, COMRef, COMFit = pyEFP.SolveKabsch(efpmol, f.subMol, f.efpcore, f.molcore)
f.rotat = [rot, COMRef, COMFit]
# deleting extra atoms
complete = 0
while (complete == 0):
complete = 1
for at in openbabel.OBMolAtomIter(efpmol):
extra = 0
if at.GetIndex()+1 not in f.efpcore:
extra = 1
for at1 in openbabel.OBAtomAtomIter(at):
if (at1.GetIndex()+1 in f.efpcore) and (at.GetAtomicNum() == 1):
extra = 0
break
if (extra == 1):
efpmol.DeleteAtom(at)
complete = 0
pyEFP.ApplyKabsch(efpmol, COMFit, COMRef, rot)
Fitmol += efpmol
return Fitmol
def save_structfile(poltype,molstruct, structfname):
"""
Intent: Output the data in the OBMol structure to a file (such as *.xyz)
Input:
molstruct: OBMol structure
structfname: output file name
Output:
file is output to structfname
Referenced By: tor_opt_sp, compute_mm_tor_energy
Description: -
"""
strctext = os.path.splitext(structfname)[1]
tmpconv = openbabel.OBConversion()
if strctext in '.xyz':
tmpfh = open(structfname, "w")
maxidx = max(poltype.symmetryclass)
iteratom = openbabel.OBMolAtomIter(molstruct)
etab = openbabel.OBElementTable()
tmpfh.write('%6d %s\n' % (molstruct.NumAtoms(), molstruct.GetTitle()))
for ia in iteratom:
tmpfh.write( '%6d %2s %13.6f %11.6f %11.6f %5d' % (ia.GetIdx(), etab.GetSymbol(ia.GetAtomicNum()), ia.x(), ia.y(), ia.z(), poltype.prmstartidx + (maxidx - poltype.symmetryclass[ia.GetIdx() - 1])))
iteratomatom = openbabel.OBAtomAtomIter(ia)
neighbors = []
for iaa in iteratomatom:
neighbors.append(iaa.GetIdx())
neighbors = sorted(neighbors)
for iaa in neighbors:
tmpfh.write('%5d' % iaa)
tmpfh.write('\n')
else:
inFormat = openbabel.OBConversion.FormatFromExt(structfname)
tmpconv.SetOutFormat(inFormat)
return tmpconv.WriteFile(molstruct, structfname)
def make_mol_paths(mol, minpath, maxpath, smarts_flag):
alist = list()
molpath = list()
for a in openbabel.OBMolAtomIter(mol):
if a.GetAtomicNum() < 2: continue # break path for * or H atom
make_atom_paths(molpath, alist, a, a, 0, minpath, maxpath, smarts_flag)
del alist[:]
return molpath
def fromOBMol(mol, obmol):
"""
Convert a OpenBabel Mol object `obmol` to a molecular structure. Uses
`OpenBabel <http://openbabel.org/>`_ to perform the conversion.
"""
# Below are the declared variables for cythonizing the module
# cython.declare(i=cython.int)
# cython.declare(radicalElectrons=cython.int, charge=cython.int, lonePairs=cython.int)
# cython.declare(atom=Atom, atom1=Atom, atom2=Atom, bond=Bond)
mol.vertices = []
# Add hydrogen atoms to complete molecule if needed
obmol.AddHydrogens()
# TODO Chem.rdmolops.Kekulize(obmol, clearAromaticFlags=True)
# iterate through atoms in obmol
for obatom in openbabel.OBMolAtomIter(obmol):
idx = obatom.GetIdx() #openbabel idx starts at 1!
# Use atomic number as key for element
number = obatom.GetAtomicNum()
element = elements.getElement(number)
# Process charge
charge = obatom.GetFormalCharge()
obatom_multiplicity = obatom.GetSpinMultiplicity()
radicalElectrons = obatom_multiplicity - 1 if obatom_multiplicity != 0 else 0
atom = Atom(element, radicalElectrons, charge, '', 0)
mol.vertices.append(atom)
# iterate through bonds in obmol
for obbond in openbabel.OBMolBondIter(obmol):
order = 0
# Process bond type
oborder = obbond.GetBondOrder()
if oborder == 1: order = 'S'
elif oborder == 2: order = 'D'
elif oborder == 3: order = 'T'
elif obbond.IsAromatic(): order = 'B'
bond = Bond(mol.vertices[obbond.GetBeginAtomIdx() - 1],
mol.vertices[obbond.GetEndAtomIdx() - 1],
order) #python array indices start at 0
mol.addBond(bond)
# Set atom types and connectivity values
mol.updateConnectivityValues()
mol.updateAtomTypes()
mol.updateMultiplicity()
mol.updateLonePairs()
# Assume this is always true
# There are cases where 2 radicalElectrons is a singlet, but
# the triplet is often more stable,
mol.multiplicity = mol.getRadicalCount() + 1
return mol
def from_ob_mol(mol, obmol, raise_atomtype_exception=True):
"""
Convert a OpenBabel Mol object `obmol` to a molecular structure. Uses
`OpenBabel <http://openbabel.org/>`_ to perform the conversion.
"""
# Below are the declared variables for cythonizing the module
# cython.declare(i=cython.int)
# cython.declare(radical_electrons=cython.int, charge=cython.int, lone_pairs=cython.int)
# cython.declare(atom=mm.Atom, atom1=mm.Atom, atom2=mm.Atom, bond=mm.Bond)
if openbabel is None:
raise DependencyError(
'OpenBabel is not installed. Please install or use RDKit.')
mol.vertices = []
# Add hydrogen atoms to complete molecule if needed
obmol.AddHydrogens()
# TODO Chem.rdmolops.Kekulize(obmol, clearAromaticFlags=True)
# iterate through atoms in obmol
for obatom in openbabel.OBMolAtomIter(obmol):
# Use atomic number as key for element
number = obatom.GetAtomicNum()
isotope = obatom.GetIsotope()
element = elements.get_element(number, isotope or -1)
# Process charge
charge = obatom.GetFormalCharge()
obatom_multiplicity = obatom.GetSpinMultiplicity()
radical_electrons = obatom_multiplicity - 1 if obatom_multiplicity != 0 else 0
atom = mm.Atom(element, radical_electrons, charge, '', 0)
mol.vertices.append(atom)
# iterate through bonds in obmol
for obbond in openbabel.OBMolBondIter(obmol):
# Process bond type
oborder = obbond.GetBondOrder()
if oborder not in [1, 2, 3, 4] and obbond.IsAromatic():
oborder = 1.5
bond = mm.Bond(mol.vertices[obbond.GetBeginAtomIdx() - 1],
mol.vertices[obbond.GetEndAtomIdx() - 1],
oborder) # python array indices start at 0
mol.add_bond(bond)
# Set atom types and connectivity values
mol.update_connectivity_values()
mol.update_atomtypes(log_species=True,
raise_exception=raise_atomtype_exception)
mol.update_multiplicity()
mol.identify_ring_membership()
# Assume this is always true
# There are cases where 2 radical_electrons is a singlet, but
# the triplet is often more stable,
mol.multiplicity = mol.get_radical_count() + 1
return mol
def testIterators(self):
"""Basic check that at least two iterators are working"""
mol = pybel.readstring("smi", "c1ccccc1C(=O)Cl")
atoms = list(ob.OBMolAtomIter(mol.OBMol))
self.assertEqual(len(atoms), 9)
elements = [atom.GetAtomicNum() for atom in atoms]
self.assertEqual(elements, [6, 6, 6, 6, 6, 6, 6, 8, 17])
bonds = list(ob.OBMolBondIter(mol.OBMol))
self.assertEqual(len(bonds), 9)
def minimization(self, options=[]):
temp1 = tempfile.NamedTemporaryFile(suffix='.pdb')
temp2 = tempfile.NamedTemporaryFile(suffix='.pdb')
bufMol = ob.OBMol()
obConversion = ob.OBConversion()
obConversion.SetInAndOutFormats('pdb', 'pdb')
obConversion.WriteFile(self.molecule, temp1.name)
popenargs = ['obminimize'] + options + [temp1.name]
subprocess.run(popenargs, stdout=temp2)
obConversion.ReadFile(bufMol, temp2.name)
temp1.close()
temp2.close()
# Update coordinates
pairs = zip(ob.OBMolAtomIter(self.molecule), ob.OBMolAtomIter(bufMol))
for (ai, aj) in pairs:
x, y, z = aj.x(), aj.y(), aj.z()
ai.SetVector(x, y, z)
def get_coordinates(self):
if self._coords is None:
self._coords = numpy.zeros((self.natoms, 3))
for at in openbabel.OBMolAtomIter(self.obmol):
i = at.GetIdx() - 1
self._coords[i, 0] = at.GetX()
self._coords[i, 1] = at.GetY()
self._coords[i, 2] = at.GetZ()
return self._coords
def get_binder_atom(binder, smarts):
log.debug("Counting atoms to find the binder atom...")
i = 1
for atom in openbabel.OBMolAtomIter(binder):
log.debug(i)
i += 1
if atom.MatchesSMARTS(smarts):
log.debug("found ya!")
return atom
def skelCN_groups(self, res):
groupnames = [
'NCA0', 'HA0', 'CO0', 'H+1', 'NCA+1', 'HA+1', 'CHB+1', 'CO+1',
'H+2', 'NCA+2', 'R'
]
groups = []
for k in groupnames:
groups.append([])
for at in openbabel.OBMolAtomIter(self.obmol):
attype = at.GetResidue().GetAtomID(at)
attype = attype.strip()
rnum = at.GetResidue().GetNum()
if attype in ['1HB', '2HB', '3HB', 'HB1', 'HB2', 'HB3']:
attype = 'HB'
if rnum == res:
if attype in ['C', 'O']:
attype = 'CO0'
elif attype in ['N', 'CA']:
attype = 'NCA0'
elif attype in ['HA']:
attype = 'HA0'
else:
attype = 'R'
elif rnum == res + 1:
if attype in ['H']:
attype = 'H+1'
elif attype in ['C', 'O']:
attype = 'CO+1'
elif attype in ['N', 'CA']:
attype = 'NCA+1'
elif attype in ['HA']:
attype = 'HA+1'
elif attype in ['CB', 'HB']:
attype = 'CHB+1'
elif rnum == res + 2:
if attype in ['H']:
attype = 'H+2'
elif attype in ['N', 'CA']:
attype = 'NCA+2'
else:
attype = 'R'
else:
attype = 'R'
try:
ind = groupnames.index(attype)
except ValueError:
print attype
raise
groups[ind].append(at.GetIdx() - 1)
return groups, groupnames
def _compute_all_sybyl(system_ob, indices=None):
sybyl_dict = {}
for atom in ob.OBMolAtomIter(system_ob):
if indices is not None and atom.GetIndex() not in indices:
continue
sybyl_dict[atom.GetIndex()] = atom.GetType()
return (sybyl_dict)
def chain_ffopt(ff, mol, frozenats):
### FORCE FIELD OPTIMIZATION ##
# INPUT
# - ff: force field to use, available MMFF94, UFF< Ghemical, GAFF
# - mol: mol3D to be ff optimized
# - connected: indices of connection atoms to metal
# - constopt: flag for constrained optimization
# OUTPUT
# - mol: force field optimized mol3D
metals = range(21, 31) + range(39, 49) + range(72, 81)
### check requested force field
ffav = 'mmff94, uff, ghemical, gaff, mmff94s' # force fields
if ff.lower() not in ffav:
print 'Requested force field not available. Defaulting to MMFF94'
ff = 'mmff94'
### convert mol3D to OBMol via xyz file, because AFTER/END option have coordinates
backup_mol = mol3D()
backup_mol.copymol3D(mol)
# print('bck ' + str(backup_mol.getAtom(0).coords()))
# print('mol_ibf ' + str(mol.getAtom(0).coords()))
mol.convert2OBMol()
### initialize constraints
constr = openbabel.OBFFConstraints()
### openbabel indexing starts at 1 ### !!!
# convert metals to carbons for FF
indmtls = []
mtlsnums = []
for iiat, atom in enumerate(openbabel.OBMolAtomIter(OBMol)):
if atom.atomicnum in metals:
indmtls.append(iiat)
mtlsnums.append(atom.GetAtomicNum())
atom.OBAtom.SetAtomicNum(19)
for cat in frozenats:
constr.AddAtomConstraint(cat + 1) # indexing babel
### set up forcefield
forcefield = openbabel.OBForceField.FindForceField(ff)
OBMol = mol.OBMol
forcefield.Setup(obmol, constr)
## force field optimize structure
forcefield.ConjugateGradients(2500)
forcefield.GetCoordinates(OBMol)
mol.OBMol = OBMol
# reset atomic number to metal
for i, iiat in enumerate(indmtls):
mol.OBMol.GetAtomById(iiat).SetAtomicNum(mtlsnums[i])
mol.convert2mol3D()
en = forcefield.Energy()
# print(str(mol.OBmol.atoms[1].OBAtom.GetVector().GetZ()))
# print(str(forcefield.Validate()))
# print('mol_af ' + str(mol.getAtom(0).coords()))
# print('ff delta = ' + str(backup_mol.rmsd(mol)))
del forcefield, constr, OBMol
return mol, en
def pymatgen_mol(self):
"""
Returns pymatgen Molecule object.
"""
sp = []
coords = []
for atom in ob.OBMolAtomIter(self._obmol):
sp.append(atom.GetAtomicNum())
coords.append([atom.GetX(), atom.GetY(), atom.GetZ()])
return Molecule(sp, coords)
