A strategy to generate random molecules for the hypothesis testing framework. It uses a collection of fragments generated from the ChEMBL database to construct plausible molecular graphs. The fragments were mined using the BRICS method.
pip install -U hypothesis-rdkit
# or
conda install -c conda-forge hypothesis-rdkitThe module hypothesis-rdkit provides a strategy for generating RDKit
molecules. During the installation of the package, this strategy is linked to the
rdkit.Chem.Mol type:
from hypothesis import given
from rdkit.Chem import Mol
@given(...)
def test_molecule_method(mol : Mol):
# mol is a randomly generated molecule
assert mol.GetNumAtoms() > 0You can use the mols strategy directly for further customization:
from hypothesis import given
from hypothesis_rdkit import mols
from rdkit.Chem import GetMolFrags, Mol
from rdkit.Chem.rdMolDescriptors import CalcNumRotatableBonds
@given(mols(n_connected_components=2, max_rotatable_bonds=5, n_conformers=10))
def test_molecule_mixtures(mol : Mol):
frags = GetMolFrags(mol, asMols=True)
assert len(frags) == 2
for frag in frags:
assert CalcNumRotatableBonds(frag) <= 5
assert mol.GetNumConformers() <= 10There are also strategies to generate molecules in SMILES, mol block or InChI
representation accepting the same parameters as mols:
from hypothesis import given
from hypothesis_rdkit import smiles, mol_blocks, inchis
from rdkit.Chem import MolFromSmiles, MolFromMolBlock, MolFromInchi
@given(smiles())
def test_smiles(smiles : str):
mol = MolFromSmiles(smiles)
assert mol is not None and mol.GetNumAtoms() > 0
@given(mol_blocks())
def test_mol_block(mol_block : str):
mol = MolFromMolBlock(mol_block)
assert mol is not None and mol.GetNumAtoms() > 0
@given(inchis())
def test_inchi(inchi : str):
mol = MolFromInchi(inchi)
assert mol is not None and mol.GetNumAtoms() > 0All fragment files are generated during a test run (pytest) in the user data
directory. On Linux, this is ~/.local/share/hypothesis_rdkit/{version}/.
