def extract_side_chains(mol, remove_duplicates=False, mark='[*]'):
""" Extract side chains from a smiles string. Core is handled as Murcko scaffold.
:param mol: {str} smiles string of a molecule.
:param remove_duplicates: {bool} Keep or remove duplicates.
:param mark: character to mark attachment points.
:return: smiles strings of side chains in a list, attachment points replaced by [R].
"""
pos = range(0, 20)
set_pos = ['[' + str(x) + '*]' for x in pos]
m1 = MolFromSmiles(mol)
try:
core = MurckoScaffold.GetScaffoldForMol(m1)
side_chain = ReplaceCore(m1, core)
smi = MolToSmiles(side_chain, isomericSmiles=True
) # isomericSmiles adds a number to the dummy atoms.
except:
return list()
for i in pos:
smi = smi.replace(''.join(set_pos[i]), mark)
if remove_duplicates:
return list(set(smi.split('.')))
else:
return smi.split('.')
def impsmiles(self, m, keepiso=False):
"""make implicit smiles from molecule"""
import re
#smi = MolToSmiles(m, canonical=True, isomericSmiles=True)
smi = MolToSmiles(m, canonical=False, isomericSmiles=keepiso)
if keepiso == False:
smi = smi.replace('([H])', '')
smi = smi.replace('[H]', '')
mol = MolFromSmiles(smi)
outsmi = ""
iatom = 0
atoms = mol.GetAtoms()
#parts = re.split("([\d\(\)\+\[\]-=#:])",smi);
parts = re.split("(\[.*?\]|Cl|Br|F|I|B|C|c|N|n|O|o|S|s|P|p)", smi)
#return [a.GetSymbol() for a in atoms]
for p in parts:
if len(p) == 0:
pass
elif p.isalpha():
hcount = atoms[iatom].GetImplicitValence()
if hcount == 0: outsmi += p
elif hcount == 1: outsmi += "[%sH]" % p
else: outsmi += "[%sH%d]" % (p, hcount)
iatom += 1
elif p.startswith("["):
hcount = atoms[iatom].GetNumImplicitHs()
if hcount == 0: outsmi += p
elif hcount == 1:
outsmi += "[%sH%d]" % (atoms[iatom].GetSymbol(),
atoms[iatom].GetFormalCharge())
else:
outsmi += "[%sH%d%+d]" % (atoms[iatom].GetSymbol(), hcount,
atoms[iatom].GetFormalCharge())
iatom += 1
else:
outsmi += p
return outsmi
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 extract_murcko_scaffolds_marked(mol, mark='[*]'):
""" Extract Bemis-Murcko scaffolds from a smile string.
:param mol: {str} smiles string of a molecule.
:param mark: character to mark attachment points.
:return: smiles string of a scaffold, side chains replaced with [R].
"""
pos = range(0, 20)
set_pos = ['[' + str(x) + '*]' for x in pos]
m1 = MolFromSmiles(mol)
try:
core = MurckoScaffold.GetScaffoldForMol(m1)
tmp = ReplaceSidechains(m1, core)
smi = MolToSmiles(tmp, isomericSmiles=True
) # isomericSmiles adds a number to the dummy atoms.
except:
return ''
for i in pos:
smi = smi.replace(''.join(set_pos[i]), mark)
return smi
