def align(self, smiles):
if isinstance(smiles, str):
amol = Chem.MolFromSmiles(smiles)
add_coords(amol)
if self.inplace:
self.add_mols()
self.data["Mol"].apply(
lambda x: Chem.GenerateDepictionMatching2DStructure(x, amol))
else:
result = self.add_mols()
result.data["Mol"].apply(
lambda x: Chem.GenerateDepictionMatching2DStructure(x, amol))
return result
def _apply(x):
if show_prog:
pb.inc()
mol = self.mol_method(x)
if mol:
add_coords(mol, force=force)
def read_sdf(fn, store_mol_as="Mol_b64", gen2d=False):
"""Create a MolFrame instance from an SD file (can be gzipped (fn ends with ``.gz``)).
Arguments:
store_mol_as: "Mol_b64" or "Smiles" """
if store_mol_as not in ["Mol_b64", "Smiles"]:
print("* Mols are stored as Mol_b64")
store_mol_as = "Mol_b64"
d = {}
ctr = 0
first_mol = True
if not isinstance(fn, list):
fn = [fn]
for f in fn:
do_close = True
if isinstance(f, str):
if f.endswith(".gz"):
file_obj = gzip.open(f, mode="rb")
else:
file_obj = open(f, "rb")
else:
file_obj = f
do_close = False
reader = Chem.ForwardSDMolSupplier(file_obj)
for mol in reader:
ctr += 1
if not mol:
print(ctr, file=sys.stderr)
continue
if first_mol:
first_mol = False
for prop in mol.GetPropNames():
if prop in [store_mol_as, "order"]:
continue
d[prop] = []
d_keys = set(d.keys())
d[store_mol_as] = []
mol_props = set()
for prop in mol.GetPropNames():
if prop in d_keys:
mol_props.add(prop)
d[prop].append(get_value(mol.GetProp(prop)))
mol.ClearProp(prop)
# append NAN to the missing props that were not in the mol:
missing_props = d_keys - mol_props
for prop in missing_props:
d[prop].append(np.nan)
if gen2d:
add_coords(mol, force=True)
if store_mol_as == "Smiles":
smi = Chem.MolToSmiles(mol)
d["Smiles"].append(smi)
else:
mol_b64 = b64.b64encode(pickle.dumps(mol)).decode()
d["Mol_b64"].append(mol_b64)
if do_close:
file_obj.close()
for k in d.keys():
print(len(d[k]))
result = MolFrame()
result.data = pd.DataFrame(d)
print_log(result.data, "read SDF")
return result