def apply_to_mol(self, new_col_name, lambda_func):
"""Applies a func to the Mol object, which is generated on-the-fly, if necessary.
Displays a progress bar for longer operations.
Returns a new copy or modifies inplace, depending on self.inplace."""
show_prog, data_len = self.which_progress(min_len=2000)
self.find_mol_col()
if show_prog is not None:
ctr = nbt.ProgCtr()
def _apply(x):
if show_prog is not None:
ctr.inc()
if ctr() % 500 == 0:
print(" - processed: {:8d}\r".format(ctr()), end="")
sys.stdout.flush()
mol = self.mol_method(x)
if not mol:
return pd.np.nan
return lambda_func(mol)
if self.inplace:
self.data[new_col_name] = self.data[self.use_col].apply(_apply)
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
else:
result = self.new()
result.data = self.data
result.data[new_col_name] = result.data[self.use_col].apply(_apply)
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
return result
def check_2d_coords(self, force=False):
"""Generates 2D coordinates if necessary.
Requires the Mol object to be present (use add_mols() ).
Always operates inplace."""
show_prog, data_len = self.which_progress(min_len=1000)
if show_prog is not None:
ctr = nbt.ProgCtr()
self.find_mol_col()
def _apply(x):
if show_prog is not None:
ctr.inc()
if ctr() % 100 == 0:
print(" - processed: {:8d}\r".format(ctr()), end="")
sys.stdout.flush()
mol = self.mol_method(x)
if mol:
check_2d_coords(mol, force=force)
if self.inplace:
self.data[self.use_col].apply(_apply)
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
else:
result = self.data[self.use_col].apply(_apply)
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
return result
def apply_to_col(self, col_name, new_col_name, lambda_func):
"""Applies a func to a column in the MolFrame.
A wrapper around pd.apply to enable progress bars.
Returns a new copy or modifies inplace, depending on self.inplace."""
show_prog, data_len = self.which_progress(min_len=5000)
if show_prog is not None:
ctr = nbt.ProgCtr()
def _apply(x):
if show_prog is not None:
ctr.inc()
if ctr() % 500 == 0:
print(" - processed: {:8d}\r".format(ctr()), end="")
sys.stdout.flush()
return lambda_func(x)
if self.inplace:
self.data[new_col_name] = self.data[col_name].apply(_apply)
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
else:
result = self.new()
result.data = self.data
result.data[new_col_name] = result.data[col_name].apply(_apply)
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
return result
def mol_filter(self, query, add_h=False):
"""Substructure filter. Returns a new MolFrame instance.
``query`` has to be a Smiles string."""
show_prog, data_len = self.which_progress(min_len=5000)
if show_prog is not None:
ctr = nbt.ProgCtr()
query_mol = Chem.MolFromSmiles(query)
if not query_mol:
raise ValueError("Could not generate query mol.")
if "[H]" in query or "#1" in query:
add_h = True
print("> explicit hydrogens turned on (add_h = True)")
res_l = []
self.find_mol_col()
for _, rec in self.data.iterrows():
if show_prog is not None:
ctr.inc()
if ctr() % 1000 == 0:
print(" - processed: {:8d}\r".format(ctr()), end="")
sys.stdout.flush()
mol = self.mol_method(rec[self.use_col])
if not mol: continue
hit = False
if add_h:
mol_with_h = Chem.AddHs(mol)
if mol_with_h.HasSubstructMatch(query_mol):
hit = True
else:
if mol.HasSubstructMatch(query_mol):
hit = True
if hit:
res_l.append(rec)
result = self.new()
result.data = pd.DataFrame(res_l)
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
print_log(result.data, "mol_filter")
return result
def sim_filter(self, query, cutoff=0.75):
"""Similarity filter. Returns a new MolFrame instance.
Add a suitable fingerprint once with addf_fps(),
then give a reference molecule or a SMILES string as query."""
if len(self.fp_name) == 0 or self.fp_col not in self.data.columns:
raise KeyError(
"No fingerprints found. Please generate them first with add_fp()."
)
show_prog, data_len = self.which_progress(min_len=5000)
if show_prog is not None:
ctr = nbt.ProgCtr()
if isinstance(query, str):
query_mol = Chem.MolFromSmiles(query)
else:
query_mol = deepcopy(query)
if not query_mol:
raise ValueError("Could not generate query mol.")
fp_method = FPDICT[self.fp_name]
query_fp = fp_method(query_mol)
res_l = []
for _, rec in self.data.iterrows():
if show_prog is not None:
ctr.inc()
if ctr() % 1000 == 0:
print(" - processed: {:8d}\r".format(ctr()), end="")
sys.stdout.flush()
mol_fp = pickle.loads(b64.b64decode(rec[self.fp_col]))
sim = DataStructs.TanimotoSimilarity(query_fp, mol_fp)
if sim >= cutoff:
rec["Sim"] = sim
res_l.append(rec)
result = self.new()
result.data = pd.DataFrame(res_l)
print_log(result.data, "sim_filter")
if show_prog is not None:
print(" - processed: {:8d} done.".format(ctr()))
return result