def test_is_valid_inchikey():
"""Test if strings are correctly classified."""
inchikeys_true = ["XYLJNLCSTIOKRM-UHFFFAOYSA-N"]
inchikeys_false = [
"XYLJNLCSTIOKRM-UHFFFAOYSA", "XYLJNLCSTIOKRMRUHFFFAOYSASN",
"XYLJNLCSTIOKR-MUHFFFAOYSA-N", "XYLJNLCSTIOKRM-UHFFFAOYSA-NN",
"Brcc(NC2=NCN2)-ccc3nccnc1-3", "2YLJNLCSTIOKRM-UHFFFAOYSA-N",
"XYLJNLCSTIOKRM-aaaaaaaaaa-a"
]
for inchikey in inchikeys_true:
assert is_valid_inchikey(inchikey), "Expected inchikey is True."
for inchikey in inchikeys_false:
assert not is_valid_inchikey(inchikey), "Expected inchikey is False."
def test_is_valid_inchikey_none_input():
"""Test None entry."""
assert not is_valid_inchikey(None), "Expected None entry to give False."
def pubchem_metadata_lookup(spectrum_in,
name_search_depth=10,
formula_search=False,
min_formula_length=6,
formula_search_depth=25,
verbose=1):
"""
Parameters
----------
spectrum_in
Matchms type spectrum as input.
name_search_depth: int
How many of the most relevant name matches to explore deeper. Default = 10.
"""
if spectrum_in is None:
return None
spectrum = spectrum_in.clone()
if is_valid_inchikey(spectrum.get("inchikey")):
return spectrum
def _plausible_name(compound_name):
return (isinstance(compound_name, str) and len(compound_name) > 4)
compound_name = spectrum.get("compound_name")
if not _plausible_name(compound_name):
return spectrum
# Start pubchem search
inchi = spectrum.get("inchi")
parent_mass = spectrum.get("parent_mass")
if isinstance(parent_mass, np.ndarray):
parent_mass = parent_mass[0]
formula = spectrum.get("formula")
# 1) Search for matching compound name
results_pubchem = pubchem_name_search(compound_name,
name_search_depth=name_search_depth,
verbose=verbose)
if len(results_pubchem) > 0:
# 1a) Search for matching inchi
if likely_has_inchi(inchi):
inchi_pubchem, inchikey_pubchem, smiles_pubchem = find_pubchem_inchi_match(
results_pubchem, inchi, verbose=verbose)
# 1b) Search for matching mass
if not likely_has_inchi(inchi) or inchikey_pubchem is None:
inchi_pubchem, inchikey_pubchem, smiles_pubchem = find_pubchem_mass_match(
results_pubchem, parent_mass, verbose=verbose)
if inchikey_pubchem is not None and inchi_pubchem is not None:
logging.info("Matching compound name: %s", compound_name)
if verbose >= 1:
print(f"Matching compound name: {compound_name}")
spectrum.set("inchikey", inchikey_pubchem)
spectrum.set("inchi", inchi_pubchem)
spectrum.set("smiles", smiles_pubchem)
return spectrum
elif verbose >= 2:
print(f"No matches found for compound name: {compound_name}")
# 2) Search for matching formula
if formula_search and formula and len(formula) >= min_formula_length:
results_pubchem = pubchem_formula_search(
formula,
formula_search_depth=formula_search_depth,
verbose=verbose)
if len(results_pubchem) > 0:
# 2a) Search for matching inchi
if likely_has_inchi(inchi):
inchi_pubchem, inchikey_pubchem, smiles_pubchem = find_pubchem_inchi_match(
results_pubchem, inchi)
# 2b) Search for matching mass
if inchikey_pubchem is None:
inchi_pubchem, inchikey_pubchem, smiles_pubchem = find_pubchem_mass_match(
results_pubchem, parent_mass)
if inchikey_pubchem is not None and inchi_pubchem is not None:
logging.info("Matching formula: %s", formula)
if verbose >= 1:
print(f"Matching formula: {formula}")
spectrum.set("inchikey", inchikey_pubchem)
spectrum.set("inchi", inchi_pubchem)
spectrum.set("smiles", smiles_pubchem)
return spectrum
elif verbose >= 2:
print(f"No matches found for formula: {formula}")
return spectrum