def _GetSMILES(mol, idxlist):
tmol = mol.__copy__() #(t)emporary
tmol = RWMol(tmol)
for AtomIdx in xrange(tmol.GetNumAtoms() - 1, -1, -1):
if AtomIdx not in idxlist:
tmol.RemoveAtom(AtomIdx)
return Chem.MolToSmiles(tmol)
def to_rdkit_molecule(data):
"""
MoleculeContainer to RDKit molecule object converter
"""
mol = RWMol()
conf = Conformer()
mapping = {}
is_3d = False
for n, a in data.atoms():
ra = Atom(a.number)
ra.SetAtomMapNum(n)
if a.charge:
ra.SetFormalCharge(a.charge)
if a.isotope != a.common_isotope:
ra.SetIsotope(a.isotope)
if a.radical:
ra.SetNumRadicalElectrons(a.radical)
mapping[n] = m = mol.AddAtom(ra)
conf.SetAtomPosition(m, (a.x, a.y, a.z))
if a.z:
is_3d = True
if not is_3d:
conf.Set3D(False)
for n, m, b in data.bonds():
mol.AddBond(mapping[n], mapping[m], _bond_map[b.order])
mol.AddConformer(conf)
SanitizeMol(mol)
return mol
def join_overclose(self, mol: Chem.RWMol, to_check, cutoff=2.2): # was 1.8
"""
Cutoff is adapted to element.
:param mol:
:param to_check: list of atoms indices that need joining (but not to each other)
:param cutoff: CC bond
:return:
"""
pt = Chem.GetPeriodicTable()
dm = Chem.Get3DDistanceMatrix(mol)
for i in to_check:
atom_i = mol.GetAtomWithIdx(i)
for j, atom_j in enumerate(mol.GetAtoms()):
# calculate cutoff if not C-C
if atom_i.GetSymbol() == '*' or atom_j.GetSymbol() == '*':
ij_cutoff = cutoff
elif atom_i.GetSymbol() == 'C' and atom_j.GetSymbol() == 'C':
ij_cutoff = cutoff
else:
ij_cutoff = cutoff - 1.36 + sum([pt.GetRcovalent(atom.GetAtomicNum()) for atom in (atom_i, atom_j)])
# determine if to join
if i == j or j in to_check:
continue
elif dm[i, j] > ij_cutoff:
continue
else:
self._add_bond_if_possible(mol, atom_i, atom_j)
def remove_exocyclic_attachments(mol):
"""
Remove exocyclic and exolinker attachments from
a molecule.
Parameters
----------
mol : rdkit.Chem.rdchem.Mol
Returns
-------
rdkit.Chem.rdchem.Mol
Molecule with exocyclic/exolinker attachments
removed.
"""
edit = RWMol(mol)
remove_atoms = set()
for atom in edit.GetAtoms():
degree = atom.GetDegree()
if degree == 1:
bond = atom.GetBonds()[0]
if bond.GetBondTypeAsDouble() == 2.0:
nbr = bond.GetOtherAtom(atom)
hcount = nbr.GetTotalNumHs()
nbr.SetNumExplicitHs(hcount + 2)
nbr.SetNoImplicit(True)
remove_atoms.add(atom.GetIdx())
for aix in sorted(remove_atoms, reverse=True):
edit.RemoveAtom(aix)
rdmolops.AssignRadicals(edit)
GetSymmSSSR(edit)
return edit.GetMol()
def _delete_marked(self, mol: Chem.RWMol):
morituri = list(
reversed(
mol.GetAtomsMatchingQuery(
Chem.rdqueries.HasPropQueryAtom('DELETE'))))
for atom in morituri:
mol.RemoveAtom(atom.GetIdx())
def get_ring_removals(smi):
rw_mol = RWMol(Chem.MolFromSmiles(smi))
rings = rw_mol.GetRingInfo().AtomRings()
out_mols = {}
for ring in rings:
new_mol = Chem.MolFromSmiles(smi)
for atom in ring:
new_mol.GetAtomWithIdx(atom).SetAtomicNum(0)
Chem.DeleteSubstructs(new_mol, Chem.MolFromSmarts("[#0]"))
Chem.GetMolFrags(new_mol)
out_mols[Chem.MolToSmiles(new_mol, isomericSmiles=True)] = ring
return out_mols
def _prevent_two_bonds_on_dummy(self, mol: Chem.RWMol):
for atom in mol.GetAtoms():
if atom.GetSymbol() != '*':
pass
elif len(atom.GetNeighbors()) <= 1:
pass
elif len(atom.GetNeighbors()) >= 2:
neighs = atom.GetNeighbors()
for second in neighs[1:]:
self._absorb(mol, atom.GetIdx(), second.GetIdx())
mol.RemoveAtom(second.GetIdx())
self._prevent_two_bonds_on_dummy(mol)
break
def apply(self, mol: RWMol) -> RWMol:
num_atoms = mol.GetNumAtoms()
if self.detach:
for i, a in enumerate(mol.GetAtoms()):
m = a.GetAtomMapNum()
if m == self.atom_map2:
for bond in a.GetBonds():
mol.RemoveBond(bond.GetBeginAtomIdx(),
bond.GetEndAtomIdx())
mol.RemoveAtom(i)
num_atoms -= 1
break
atom_ind = get_atom_ind(mol, self.atom_map1)
b_type = rdchem.BondType.values[self.bond_type]
b_stereo = rdchem.BondStereo.values[self.bond_stereo]
old_atom = mol.GetAtomWithIdx(atom_ind)
if old_atom.HasProp('in_reactant'):
self.new_a.SetBoolProp('in_reactant',
old_atom.GetBoolProp('in_reactant'))
if old_atom.HasProp('mol_id'):
self.new_a.SetIntProp('mol_id', old_atom.GetIntProp('mol_id'))
mol.AddAtom(self.new_a)
new_atom_ind = num_atoms
bond_ind = mol.AddBond(atom_ind, new_atom_ind, order=b_type) - 1
new_bond = mol.GetBondWithIdx(bond_ind)
new_bond.SetStereo(b_stereo)
new_bond.SetBoolProp('is_edited', True)
return mol
def _prevent_bridge_ring(self, mol: Chem.RWMol, examplar: Tuple[int]):
## This is really
# examplar is ring
ringatoms = self._get_ring_info(mol) #GetRingInfo().AtomRings()
ringatoms = [ring for ring in ringatoms if set(ring).intersection(examplar)]
ring_idx = list(range(len(ringatoms)))
shared_count = {}
for ra, rb in itertools.combinations(ring_idx, r=2):
shared_count[(ra, rb)] = len(set(ringatoms[ra]).intersection(set(ringatoms[rb])))
if len(shared_count) == 0:
return mol
ra, rb = list(shared_count.keys())[0]
shared = list(set(ringatoms[ra]).intersection(ringatoms[rb]))
pairs = [(a, b) for a, b in itertools.combinations(shared, r=2) if mol.GetBondBetweenAtoms(a, b) is not None]
c = Counter([i for pair in pairs for i in pair])
ring_A, ring_B = ringatoms[ra], ringatoms[rb]
small, big = sorted([ring_A, ring_B], key=lambda ring: len(ring))
inners = [i for i in c if c[i] > 1]
x = list(set(shared).difference(inners))
if len(x) != 2:
log.critical(f'This is impossible. {ringatoms} share {shared} with {inners} in the inside and {x} on the edge?')
return mol
a, b = x
if len(big) > 6:
log.warning(f'Removing {len(inners)} bridging atoms and replacing with fused ring')
# bond the vertices
bt = Chem.BondType.SINGLE # ???
if mol.GetBondBetweenAtoms(a, b) is None:
mol.AddBond(a, b, bt)
else:
log.warning('This is really odd! Why is there a bond already??')
# remove the middle atoms.
for i in sorted(inners, reverse=True):
mol.RemoveAtom(i)
else:
log.warning(f'Shriking the smaller ring to change from bridged to fused.')
# get the neighbour in the small atom to a vertex.
neighs = [neigh for neigh in mol.GetAtomWithIdx(a).GetNeighbors() if
neigh.GetIdx() not in shared and neigh.GetIdx() in small]
neigh = sorted(neighs, key=lambda atom: atom.GetSymbol() != 'C')[0]
bt = mol.GetBondBetweenAtoms(a, neigh.GetIdx()).GetBondType()
mol.RemoveBond(a, neigh.GetIdx())
new_neigh = [neigh for neigh in mol.GetAtomWithIdx(a).GetNeighbors() if neigh.GetIdx() in shared][0]
mol.AddBond(neigh.GetIdx(), new_neigh.GetIdx(), bt)
neigh.SetBoolProp('_Novel', True)
new_neigh.SetBoolProp('_Novel', True)
mol.GetAtomWithIdx(a).SetBoolProp('_Novel', True)
return mol
def join_rings(self, mol: Chem.RWMol, cutoff=1.8):
# special case: x0749. bond between two rings
# namely bonds are added to non-ring atoms. so in the case of bonded rings this is required.
rings = self._get_ring_info(mol)
dm = Chem.Get3DDistanceMatrix(mol)
for ringA, ringB in itertools.combinations(rings, 2):
if not self._are_rings_bonded(mol, ringA, ringB):
mini = np.take(dm, ringA, 0)
mini = np.take(mini, ringB, 1)
d = np.nanmin(mini)
if d < cutoff:
p = np.where(mini == d)
f = ringA[int(p[0][0])]
s = ringB[int(p[1][0])]
#mol.AddBond(f, s, Chem.BondType.SINGLE)
self._add_bond_if_possible(mol, mol.GetAtomWithIdx(f), mol.GetAtomWithIdx(s))
def step(self, action_ob):
"""
Used to perform actions on the current molecule in the environment
:param action_ob: The action to be taken on the current molecule
:type action_ob: Action
:return: Information about the resulting molecule
:rtype: Observation
"""
action = action_ob.action_c.lower()
position = action_ob.pos
mol = action_ob.mol
query = action_ob.query
if (isinstance(action_ob.query, np.ndarray)):
self._queryStep(action, query)
else:
self._simpleStep(action, position, mol)
self.current_molecule = RWMol(Chem.MolFromSmiles(self._listToSmiles()))
self.datacapture.processing()
self.obs.update(self.current_molecule)
self.mol_Steps.append(self.current_molecule)
legend = str(len(self.mol_Steps)) + ". " + Chem.MolToSmiles(
self.current_molecule)
self.smiles.append(legend)
if os.environ.get('DISPLAY', '') != '':
img = Draw.MolToImage(self.current_molecule,
size=(300, 300),
kekulize=True,
wedgeBonds=True)
self.gui.update(img)
return self.obs
def __init__(self):
"""
This is the constructor
"""
super().__init__()
default_smile = 'C'
self.current_molecule = RWMol(Chem.MolFromSmiles(default_smile))
self.obs = Observation(self.current_molecule)
self.molecule_list = [Mol_Feature(default_smile)]
self.datacapture = Datacapture(self.current_molecule)
self.datacapture.processing()
self.mol_Steps = [self.current_molecule]
legend = str(len(self.mol_Steps)) + ". " + Chem.MolToSmiles(
self.current_molecule)
self.smiles = [legend]
if os.environ.get('DISPLAY',
'') != '': #check if there is a display available
self.root = Toplevel()
self.gui = Render(self.root)
img = Draw.MolToImage(self.current_molecule,
size=(300, 300),
kekulize=True,
wedgeBonds=True)
self.gui.update(img)
else:
print('No display found!')
def _prevent_two_bonds_on_dummy(self, mol: Chem.RWMol):
"""
The case '*(C)C' is seen legitimately in some warheads... but in most cases these are not.
:param mol:
:return:
"""
for atom in mol.GetAtoms():
if atom.GetSymbol() != '*':
pass
elif len(atom.GetNeighbors()) <= 1:
pass
elif len(atom.GetNeighbors()) >= 2:
self.journal.info(
f'Dummy atom (idx={atom.GetIdx()}) has {len(atom.GetNeighbors())} bonds!'
)
neighs = atom.GetNeighbors()
first = neighs[0]
for second in neighs[1:]:
rejected = second.GetIdx(
) # that will be absorbed (deleted)
keeper = first.GetIdx() # that absorbs (kept)
self._copy_bonding(mol, keeper, rejected)
self._mark_for_deletion(mol, rejected)
self._delete_marked(mol)
return self._prevent_two_bonds_on_dummy(mol)
def rd_map_from_ob(mol):
from rdkit.Chem import RWMol, Atom, BondType
rm = RWMol()
for i in range(mol.NumAtoms()):
a = mol.GetAtomById(i)
ra = Atom(a.GetAtomicNum())
rm.AddAtom(ra)
for i in range(mol.NumBonds()):
b = mol.GetBondById(i)
b.GetBeginAtom().GetId()
order = BondType.SINGLE
if b.GetBO() == 2:
order = BondType.DOUBLE
if b.GetBO() == 3:
order = BondType.TRIPLE
rm.AddBond(b.GetBeginAtom().GetId(), b.GetEndAtom().GetId(),order)#b.GetBondOrder())
return rm
def _place_between(self, mol: Chem.RWMol, a: int, b: int, aromatic=True):
oribond = mol.GetBondBetweenAtoms(a, b)
if oribond is None:
print('FAIL')
return None # fail
elif aromatic:
bt = Chem.BondType.AROMATIC
else:
bt = oribond.GetBondType()
idx = mol.AddAtom(Chem.Atom(6))
neoatom = mol.GetAtomWithIdx(idx)
atom_a = mol.GetAtomWithIdx(a)
atom_b = mol.GetAtomWithIdx(b)
if aromatic:
neoatom.SetIsAromatic(True)
atom_a.SetIsAromatic(True)
atom_b.SetIsAromatic(True)
# prevent constraints
neoatom.SetBoolProp('_Novel', True)
atom_a.SetBoolProp('_Novel', True)
atom_b.SetBoolProp('_Novel', True)
# fix position
conf = mol.GetConformer()
pos_A = conf.GetAtomPosition(a)
pos_B = conf.GetAtomPosition(b)
x = pos_A.x / 2 + pos_B.x / 2
y = pos_A.y / 2 + pos_B.y / 2
z = pos_A.z / 2 + pos_B.z / 2
conf.SetAtomPosition(idx, Point3D(x, y, z))
# fix bonds
mol.RemoveBond(a, b)
mol.AddBond(a, idx, bt)
mol.AddBond(b, idx, bt)
def complement_reaction(rxn_template):
if rxn_template.GetNumProductTemplates() != 1:
print("[ERROR] A reaction template has only one product template.")
sys.exit(1)
pro = rxn_template.GetProductTemplate(0)
rw_pro = RWMol(pro)
amaps_pro = {a.GetAtomMapNum() for a in pro.GetAtoms()}
amaps_rcts = {a.GetAtomMapNum() for rct in rxn_template.GetReactants() for a in rct.GetAtoms()}
amaps_not_in_rcts = amaps_pro.intersection(amaps_rcts)
for amap in amaps_not_in_rcts:
aidx = [a.GetIdx() for a in rw_pro.GetAtoms() if a.GetAtomMapNum() == amap][0]
rw_pro.RemoveAtom(aidx)
m = rw_pro.GetMol()
if '.' in Chem.MolToSmarts(m):
return
if (m.GetNumAtoms() == 0) or (m.GetNumAtoms() == 1 and m.GetAtomWithIdx(0).GetSymbol() in {"*", None}):
return
rxn_template.AddReactantTemplate(m)
def fragment(self, scaffold):
"""Fragment a scaffold into its next set of Murcko fragments.
Parameters
----------
scaffold : scaffoldgraph.core.Scaffold
Child scaffold to be fragmented.
Returns
-------
list
A list of parent scaffolds representing the next hierarchy.
"""
parents = []
rings = scaffold.ring_systems # ring system information
info = scaffold.rings.info
if rings.count == 1:
return []
for rix, ring in enumerate(rings):
edit = RWMol(scaffold.mol)
remove_atoms = set()
for index, atom in zip(ring.aix, ring.atoms):
if info.NumAtomRings(index) == 1 or any(
[not b.IsInRing() for b in atom.GetBonds()]):
if atom.GetDegree() > 2: # Evoke linker collection
collect_linker_atoms(edit.GetAtomWithIdx(index),
remove_atoms)
else:
remove_atoms.add(index)
else:
remove_atoms.add(index)
for aix in sorted(remove_atoms, reverse=True):
edit.RemoveAtom(aix)
for parent in get_scaffold_frags(edit):
if parent.ring_systems.count == len(rings) - 1:
parent.removed_ring_idx = rix
parents.append(parent)
return parents
def Write(self, degrees, edges, canonical=True):
if set(degrees).issubset(self.allowed):
# Define the molecule
cp = RWMol()
_ = [cp.AddAtom(Atom(self.d2atno[D])) for D in degrees]
_ = [cp.AddBond(f, t, BondType.SINGLE) for f, t in edges]
# Export as canonical SMILES or a random SMILES
if canonical:
out = MolToSmiles(cp, canonical=True)
else:
out = MolToSmiles(cp, canonical=False, doRandom=True)
# Carry out replacements
for src, dst in self.replacements:
out = out.replace(src, dst)
return out.upper()
else:
return None
def fragment(self, scaffold):
"""Fragment a scaffold into its next set of murcko fragments.
This fragmenter will not dissect fused ring systems.
Parameters
----------
scaffold (sg.core.Scaffold): scaffold to be fragmented.
Returns
-------
parents (list): a list of the next scaffold parents.
"""
parents = []
rings = scaffold.ring_systems # ring system information
info = scaffold.rings.info
if rings.count == 1:
return []
for rix, ring in enumerate(rings):
edit = RWMol(scaffold.mol)
remove_atoms = set()
for index, atom in zip(ring.aix, ring.atoms):
if info.NumAtomRings(index) == 1:
if atom.GetDegree() > 2: # Evoke linker collection
collect_linker_atoms(edit.GetAtomWithIdx(index),
remove_atoms)
else:
remove_atoms.add(index)
else:
remove_atoms.add(index)
for aix in sorted(remove_atoms, reverse=True):
edit.RemoveAtom(aix)
for parent in get_scaffold_frags(edit):
if parent.ring_systems.count == len(rings) - 1:
parent.removed_ring_idx = rix
parents.append(parent)
return parents
def to_rdkit_molecule(data: MoleculeContainer):
"""
MoleculeContainer to RDKit molecule object converter
"""
mol = RWMol()
mapping = {}
bonds = data._bonds
for n, a in data.atoms():
ra = Atom(a.atomic_number)
ra.SetAtomMapNum(n)
if a.charge:
ra.SetFormalCharge(a.charge)
if a.isotope:
ra.SetIsotope(a.isotope)
if a.is_radical:
ra.SetNumRadicalElectrons(1)
mapping[n] = mol.AddAtom(ra)
for n, m, b in data.bonds():
mol.AddBond(mapping[n], mapping[m], _bond_map[b.order])
for n in data._atoms_stereo:
ra = mol.GetAtomWithIdx(mapping[n])
env = bonds[n]
s = data._translate_tetrahedron_sign(n, [x for x in mapping if x in env])
ra.SetChiralTag(_chiral_ccw if s else _chiral_cw)
for nm, s in data._cis_trans_stereo.items():
n, m = nm
if m in bonds[n]: # cumulenes unsupported
nn, nm, *_ = data._stereo_cis_trans[nm]
b = mol.GetBondBetweenAtoms(mapping[n], mapping[m])
b.SetStereoAtoms(mapping[nn], mapping[nm])
b.SetStereo(_cis if s else _trans)
conf = Conformer()
for n, a in data.atoms():
conf.SetAtomPosition(mapping[n], (a.x, a.y, 0))
conf.Set3D(False)
mol.AddConformer(conf, assignId=True)
for c in data._conformers:
conf = Conformer()
for n, xyz in c.items():
conf.SetAtomPosition(mapping[n], xyz)
mol.AddConformer(conf, assignId=True)
SanitizeMol(mol)
AssignStereochemistry(mol, flagPossibleStereoCenters=True, force=True)
return mol
def apply(self, mol: RWMol) -> RWMol:
atom_ind = get_atom_ind(mol, self.atom_map1)
atom = mol.GetAtomWithIdx(atom_ind)
atom.SetFormalCharge(self.formal_charge)
a_chiral = rdchem.ChiralType.values[self.chiral_tag]
atom.SetChiralTag(a_chiral)
atom.SetNumExplicitHs(self.num_explicit_hs)
atom.SetIsAromatic(self.is_aromatic)
atom.SetBoolProp('is_edited', True)
return mol
def decode(v):
"""Decode a molvector into a molecule
:param v: molvector
:result rdkit.RWMol:
"""
chunksize = atom_size + bond_chunk_size
nchunks = len(v) // chunksize
m = RWMol()
bonds = {}
for i in range(nchunks):
start = i * (atom_size + bond_chunk_size)
el, c, h, b1, o1, b2, o2, b3, o3, b4, o4 = v[start:start + chunksize]
atom = Atom(el)
atom.SetFormalCharge(c)
atom.SetNumExplicitHs(h)
atom_idx = m.AddAtom(atom)
assert atom_idx == i
for b, o in ((b1, o1), (b2, o2), (b3, o3), (b4, o4)):
if o:
to_atom = atom_idx + o
bonds[tuple(sorted((atom_idx, to_atom)))] = b
for (a1, a2), btype in bonds.items():
try:
m.AddBond(a1 % m.GetNumAtoms(), a2 % m.GetNumAtoms(),
BondType.values[btype])
except:
pass
return m
def to_rdkit_molecule(data):
"""
MoleculeContainer to RDKit molecule object converter
"""
mol = RWMol()
mapping = {}
for n, a in data.atoms():
ra = Atom(a.atomic_number)
ra.SetAtomMapNum(n)
if a.charge:
ra.SetFormalCharge(a.charge)
if a.isotope:
ra.SetIsotope(a.isotope)
if a.is_radical:
ra.SetNumRadicalElectrons(1)
mapping[n] = mol.AddAtom(ra)
for n, m, b in data.bonds():
mol.AddBond(mapping[n], mapping[m], _bond_map[b.order])
conf = Conformer()
for n, a in data.atoms():
conf.SetAtomPosition(mapping[n], (a.x, a.y, 0))
conf.Set3D(False)
mol.AddConformer(conf)
for c in data._conformers:
conf = Conformer()
for n, xyz in c.items():
conf.SetAtomPosition(mapping[n], xyz)
mol.AddConformer(conf)
SanitizeMol(mol)
return mol
def convert_dict_to_mols(tot_dict):
"""
:param tot_dict:
:return:
"""
mol_list = []
for smiles in tot_dict:
# Now generate the molecules for that
mol = RWMol()
atoms = tot_dict[smiles]
print(atoms)
for atom in atoms:
atom = Atom(6)
mol.AddAtom(atom)
# for i in range(len(atoms)-1):
# mol.AddBond(i,i+1)
mol = mol.GetMol()
AllChem.EmbedMolecule(mol)
conf = mol.GetConformer()
for i, atom in enumerate(atoms):
point_3d = Point3D(atom[0], atom[1], atom[2])
conf.SetAtomPosition(i, point_3d)
mol = conf.GetOwningMol()
mol.SetProp("_Name", smiles)
mol_list.append(mol)
return mol_list
def apply(self, mol: RWMol) -> RWMol:
atom1 = get_atom_ind(mol, self.atom_map1)
atom2 = get_atom_ind(mol, self.atom_map2)
if self.bond_type is None: # delete bond
bond = mol.GetBondBetweenAtoms(atom1, atom2)
if bond is not None:
mol.RemoveBond(atom1, atom2)
else:
b_type = rdchem.BondType.values[self.bond_type]
b_stereo = rdchem.BondStereo.values[self.bond_stereo]
bond = mol.GetBondBetweenAtoms(atom1, atom2)
if bond is None: # add new bond
bond_ind = mol.AddBond(atom1, atom2, order=b_type) - 1
bond = mol.GetBondWithIdx(bond_ind)
else: # change an existing bond
bond.SetBondType(b_type)
bond.SetStereo(b_stereo)
bond.SetBoolProp('is_edited', True)
if b_type == BondType.AROMATIC:
bond.SetIsAromatic(True)
mol.GetAtomWithIdx(atom1).SetIsAromatic(True)
mol.GetAtomWithIdx(atom2).SetIsAromatic(True)
return mol
def depict(self, filename=None, ipython=False):
from rdkit.Chem.Draw import IPythonConsole
from rdkit.Chem.Draw import MolToImage
from rdkit.Chem.Draw import rdMolDraw2D
from rdkit.Chem.AllChem import EmbedMolecule
from IPython.display import SVG
from rdkit.Chem import RWMol, MolFromSmiles, Atom, BondType, ChiralType
_ = MolFromSmiles('C')
rmol = RWMol(_)
dict_old_new_idx = {}
n = 1
for a in self.atoms:
old_idx = a.GetIdx()
rmol.AddAtom(a)
dict_old_new_idx[old_idx] = n
n += 1
for a in self.enviroments:
old_idx = a.GetIdx()
a.SetChiralTag(ChiralType.CHI_UNSPECIFIED)
a.SetIsAromatic(0)
rmol.AddAtom(a)
dict_old_new_idx[old_idx] = n
n += 1
for b in self.Bonds:
rmol.AddBond(dict_old_new_idx[b.GetBeginAtomIdx()],
dict_old_new_idx[b.GetEndAtomIdx()], b.GetBondType())
for b in self.bondsenvironments:
rmol.AddBond(dict_old_new_idx[b.GetBeginAtomIdx()],
dict_old_new_idx[b.GetEndAtomIdx()], b.GetBondType())
rmol.RemoveAtom(0)
EmbedMolecule(rmol)
drawer = rdMolDraw2D.MolDraw2DSVG(400, 200)
drawer.DrawMolecule(rmol)
drawer.FinishDrawing()
svg = drawer.GetDrawingText()
if filename != None:
f = open(filename, 'w')
f.write(svg)
f.close()
if ipython:
svg = svg.replace('svg:', '')
return SVG(svg)
else:
return None
def _restore_original_bonding(self, mol: Chem.RWMol, rings) -> None:
to_be_waited_for = []
for ring in rings:
for i in range(len(ring['elements'])):
d = self._get_expansion_for_atom(ring, i)
new_i = self._get_new_index(mol, d['ori_i'], search_collapsed=False)
for old_neigh, bond in zip(d['neighbor'], d['bond']):
bt = getattr(Chem.BondType, bond)
try:
new_neigh = self._get_new_index(mol, old_neigh, search_collapsed=False)
present_bond = mol.GetBondBetweenAtoms(new_i, new_neigh)
if present_bond is None:
mol.AddBond(new_i, new_neigh, bt)
elif present_bond.GetBondType().name != bond:
if self._debug_draw:
print(
f'bond between {new_i} {new_neigh} exists already (has {present_bond.GetBondType().name} expected {bt})')
present_bond.SetBondType(bt)
else:
if self._debug_draw:
print(f'bond between {new_i} {new_neigh} exists already ' + \
f'(has {present_bond.GetBondType().name} expected {bt})')
pass
except ValueError:
if self._debug_draw:
print(f"The neighbour {old_neigh} of {d['ori_i']} with {bt} does not yet exist")
to_be_waited_for.append((new_i, old_neigh, bt))
for new_i, old_neigh, bt in to_be_waited_for:
try:
new_neigh = self._get_new_index(mol, old_neigh, name_restriction=mol.GetAtomWithIdx(new_i).GetProp('_ori_name'))
if self._debug_draw:
print(f'{old_neigh} was missing, but has appeared since as {new_neigh}')
if not mol.GetBondBetweenAtoms(new_i, new_neigh):
mol.AddBond(new_i, new_neigh, bt)
except (KeyError, ValueError) as err:
warn(str(err))
def test_step(self):
#test add-back
smile = "CC"
smile = Chem.CanonSmiles(smile)
m = Chem.CanonSmiles(self.env._listToSmiles())
self.assertEqual(m, smile)
mols = []
legends = []
mols.append(RWMol(Chem.MolFromSmiles("C")))
legends.append("1. C")
mols.append(RWMol(Chem.MolFromSmiles("CC")))
legends.append("2. CC")
#test add-front
self.action.setAction("add", pos="front", mol="C1=CC=CC=C1")
self.env.step(self.action)
smile = "CCC1=CC=CC=C1"
smile = Chem.CanonSmiles(smile)
m = Chem.CanonSmiles(self.env._listToSmiles())
self.assertEqual(m, smile)
mols.append(RWMol(Chem.MolFromSmiles("CCC1=CC=CC=C1")))
l = "3. " + self.env._listToSmiles()
legends.append(l)
#test remove-back
self.action.setAction("remove", pos="back", mol="C")
self.env.step(self.action)
smile = "C1=CC=CC=C1C"
smile = Chem.CanonSmiles(smile)
m = Chem.CanonSmiles(self.env._listToSmiles())
self.assertEqual(m, smile)
mols.append(RWMol(Chem.MolFromSmiles("CC1=CC=CC=C1")))
l = "3. " + self.env._listToSmiles()
legends.append(l)
#test remove-front
self.action.setAction("remove", pos="front", mol="C1=CC=CC=C1")
self.env.step(self.action)
smile = "C"
smile = Chem.CanonSmiles(smile)
m = Chem.CanonSmiles(self.env._listToSmiles())
self.assertEqual(m, smile)
mols.append(RWMol(Chem.MolFromSmiles("C")))
l = "3. " + self.env._listToSmiles()
legends.append(l)
#test current molecule
mol = self.env.current_molecule
self.action.setAction("add", pos="front", mol="CC")
self.env.step(self.action)
self.assertNotEqual(self.env.current_molecule, mol)
mols.append(RWMol(Chem.MolFromSmiles("CCC")))
l = "3. " + self.env._listToSmiles()
legends.append(l)
def seed(self, Smiles):
"""
Resets the environment to a specified molecule
:param Smiles: Smiles string for the molecule
:type Smiles: string
"""
self.current_molecule = RWMol(Chem.MolFromSmiles(Smiles))
self.molecule_list = [Mol_Feature(Smiles)]
self.obs = Observation(self.current_molecule)
self.mol_Steps = [self.current_molecule]
legend = str(len(self.mol_Steps)) + ". " + Chem.MolToSmiles(
self.current_molecule)
self.smiles.append(legend)
self.datacapture = Datacapture(self.current_molecule)
self.datacapture.processing()
def _minimize_rings(mol):
"""Private: Minimize rings in a scaffold.
In this process, all remaining vertices/atoms of degree two are
removed by performing an edge merging operation. The only
exception being when both vertices neighbours are connected
(i.e. we have a triangle), when edge merging would lead to the
loss of a cycle. The result is a minimum cycle topological
representation of the original molecule. This function is used
in the computation of ring topology scaffolds (Oprea).
If a ring contains a non-carbon atom, this atom is maintained.
Neighbouring ring atoms which are of the same type are merged
together into a single atom of the corresponding type.
Parameters
----------
mol : rdkit.Chem.rdchem.Mol
Returns
-------
rdkit.Chem.rdchem.RWMol
Minimum cycle topological graph.
"""
edit = RWMol(mol)
remove_atoms = set()
for atom in edit.GetAtoms():
if atom.GetDegree() == 2:
n1, n2 = atom.GetNeighbors()
n1_idx, n2_idx = n1.GetIdx(), n2.GetIdx()
connected = edit.GetBondBetweenAtoms(n1_idx, n2_idx)
if not connected and (n1.GetAtomicNum() == atom.GetAtomicNum()
or n2.GetAtomicNum() == atom.GetAtomicNum()):
a_idx = atom.GetIdx()
edit.RemoveBond(n1_idx, a_idx)
edit.RemoveBond(n2_idx, a_idx)
edit.AddBond(n1_idx, n2_idx, BondType.SINGLE)
remove_atoms.add(a_idx)
for a_idx in sorted(remove_atoms, reverse=True):
edit.RemoveAtom(a_idx)
return edit