def export_sdf(mine_db: MINE,
dir_path: str,
max_compounds: int = None) -> None:
"""Exports compounds from the database as an MDL SDF file.
Parameters
----------
mine_db : MINE
MINE object that contains the database.
dir_path : str
Directory for files.
max_compounds : int, optional
Maximum number of compounds per file, by default None.
"""
# Make sure that all compounds point to all their reactants
if not mine_db.compounds.find_one({"Product_of": {"$exists": 1}}):
mine_db.add_rxn_pointers()
print(
f"Exporting {mine_db.compounds.count()} compounds from {mine_db.name}"
" as an SDF file")
target = utils.prevent_overwrite(
os.path.join(dir_path, mine_db.name) + "_1.sdf")
# SDWriter (rdkit) writes Mol objects to SD files
writer = AllChem.SDWriter(target)
writer.SetKekulize(True)
n_files = 1
for compound in mine_db.compounds.find():
# Convert SMILES string to Mol object, replacing 'CoA' and 'R' by '*'
mol = AllChem.MolFromSmiles(compound["SMILES"], True, {
"CoA": "*",
"R": "*"
})
# if Mol object successfully generated, annotate properties
if mol:
mol.SetProp("_id", compound["_id"])
mol.SetProp("Generation", str(compound["Generation"]))
if "Reactant_in" in compound:
mol.SetProp("Reactant_in", str(compound["Reactant_in"]))
if "Product_of" in compound:
mol.SetProp("Product_of", str(compound["Product_of"]))
writer.write(mol)
# Start writing a new sdf file if the maximum (set by user) has
# been reached for the current file
if max_compounds and (writer.NumMols() >= max_compounds):
n_files += 1
target = utils.prevent_overwrite(
os.path.join(dir_path, mine_db.name) + f"_(n_files).sdf")
writer = AllChem.SmilesWriter(target)
writer.close()
def export_smiles(mine_db: MINE,
dir_path: str,
max_compounds: int = None) -> None:
"""Exports compounds from the database as a SMILES file.
Parameters
----------
mine_db : MINE
MINE object that contains the database.
dir_path : str
Directory for files.
max_compounds : int, optional
Maximum number of compounds per file, by default None.
"""
header = ["SMILES", "_id", "Generation", "Reactant_in", "Product_of"]
# Make sure that all compounds point to all their reactants
if not mine_db.compounds.find_one({"Product_of": {"$exists": 1}}):
mine_db.add_rxn_pointers()
print(
f"Exporting {mine_db.compounds.count()} compounds from {mine_db.name()}"
" as SMILES file")
target = open(
utils.prevent_overwrite(
os.path.join(dir_path, mine_db.name) + "_1.smiles"), "w")
# DictWriter allows for each key:value pair of a dictionary to be written
# on its own row (by writerow)
writer = csv.DictWriter(target, fieldnames=header, dialect="excel-tab")
n_files = 1
i = 0
for compound in mine_db.compounds.find({}, dict([(x, 1) for x in header])):
writer.writerow(compound)
i += 1
# If max compounds per file has been set by user and our number of
# compounds that we have written so far is divisible by the max number,
# then we start a new file
if max_compounds and not i % max_compounds:
n_files += 1
target = open(
utils.prevent_overwrite(
os.path.join(dir_path, mine_db.name) +
f"_{n_files}.smiles"),
"w",
)
writer = csv.DictWriter(target,
fieldnames=header,
dialect="excel-tab")