def getChemblSimilarLigandsBySmile(smi, threshold=85, returnSmiles=False):
"""
Returns a SmallMolLib object of the ligands having a similarity with a smile of at least the specified
threshold.. This molecules are retrieve from Chembl. It is possible to return also the list smiles.
Parameters
----------
smi: str
The smile
threshold: int
The threshold value to apply for the similarity search
returnSmiles: bool
If True, the list smiles is returned
Returns
-------
sm: htmd.smallmol.smallmol.SmallMol
The SmallMol object
smiles: str
The list of smiles
Example
-------
>>> _, smile = getChemblLigandByDrugName('ibuprofen', returnSmile=True) # doctest: +SKIP
>>> lib = getChemblSimilarLigandsBySmile(smile) # doctest: +SKIP
>>> lib.numMols # doctest: +SKIP
4
>>> lib, smiles = getChemblSimilarLigandsBySmile(smile, returnSmiles=True) # doctest: +SKIP
>>> len(smiles) # doctest: +SKIP
4
"""
from htmd.smallmol.smallmol import SmallMolLib, SmallMol
try:
from chembl_webresource_client.new_client import new_client
except ImportError as e:
raise ImportError(
'You need to install the chembl_webresource package to use this function. Try using `conda install '
'-c chembl chembl_webresource_client`.')
smi_list = []
similarity = new_client.similarity
results = similarity.filter(smiles=smi, similarity=threshold).only(['molecule_structures'])
results = results.all()
for r in range(len(results)):
tmp_smi = results[r]['molecule_structures']['canonical_smiles']
fragments = tmp_smi.split('.')
fragments_len = [ len(fr) for fr in fragments ]
fragment = fragments[fragments_len.index(max(fragments_len))]
if fragment not in smi_list: smi_list.append(fragment)
lib = SmallMolLib()
for smi in smi_list: lib.appendSmallMol(SmallMol(smi))
if returnSmiles:
return lib, smi_list
return lib
def test_03_appendSmallMolLib(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
lib2 = SmallMolLib(sdffile)
lib.appendSmallLib(lib2)
n_mol2_merged = lib.numMols
self.assertEqual(n_mol2_merged, SDF_N_MOLS*2, msg="The number of molecules in the SmallMolLib is not as expected."
"The two sdf were not correctly merged. ")
def test_04_appendSmallMol(self):
mol2file = os.path.join(self.dataDir, 'benzamidine.mol2')
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
sm = SmallMol(mol2file)
lib.appendSmallMol(sm)
n_mol2_append = lib.numMols
self.assertEqual(n_mol2_append, SDF_N_MOLS+1, msg="The number of molecules in the SmallMolLib is not as expected."
"The mol2 were not correctly append. ")
def test_02_writeSdf(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
sdfname = NamedTemporaryFile().name + '.sdf'
lib.writeSdf(sdfname)
sdf_exists = os.path.isfile(sdfname)
self.assertTrue(sdf_exists, msg="The sdf written was not found")
sdf = SmallMolLib(sdfname)
self.assertIsInstance(sdf, SmallMolLib, msg="The sdf written was not correctly loaded. Probably the previous"
"writing went wrong")
def test_04_loadSdffile(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
n_mols = lib.numMols
self.assertEqual(
n_mols, SDF_N_MOLS, 'Molecules not correctly loaded. '
'Expected: {}; Now: {}'.format(SDF_N_MOLS, n_mols))
def test_00_getCommonStructure(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
rd_mols = lib._mols
cms, cms_idxs, other_idxs = getMaximumCommonSubstructure(
rd_mols, returnAtomIdxs=True)
self.assertIsInstance(cms,
rdkit.Chem.rdchem.Mol,
msg="The object is not a rdkit Molecule object")
ref_cms_natoms = CMS_N_ATOMS
ref_atomidx_mol0 = CMS_AIDX_MOL0
cms_atomidx_mol0 = cms_idxs[0]
cms_natoms = len(cms_atomidx_mol0)
self.assertEqual(cms_atomidx_mol0,
ref_atomidx_mol0,
msg="The atomidx of the CMS for the mol0 are not the"
" expected ones")
self.assertEqual(
cms_natoms,
ref_cms_natoms,
msg="The number of atom in the CMS for the mol0 are not the"
" expected ones")
def test_07_depict(self):
import IPython
refimg = os.path.join(self.dataDir, 'sdf.svg')
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
img_name = NamedTemporaryFile().name + '.svg'
lib.depict(sketch=True, filename=img_name)
_img = lib.depict(sketch=True, ipython=True)
refimg_size = os.path.getsize(refimg)
sm_img_size = os.path.getsize(img_name)
self.assertIsInstance(_img, IPython.core.display.SVG, msg="The object is not an IPython image as expected")
self.assertEqual(sm_img_size, refimg_size, msg="The svg image does not have the same size of the reference")
def test_06_convertToDataFrame(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
df = lib.toDataFrame()
self.assertIsInstance(df, core.frame.DataFrame, msg="The SmallMolLib object was not correctly converted into pandas"
"DataFrame")
cols = df.columns.tolist()
ref_cols =SDF_FIELDS
self.assertEqual(cols, ref_cols, msg="The fields in the SmallMolLib object was not the expected one")
ligname_99 = df.iloc[99][0]
ref_ligname = SDF_LOC_0_99
self.assertEqual(ligname_99, ref_ligname, msg="The ligand name found is not the expected one")
def test_06_cluster_shape(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
mols = lib._mols
cl, det = cluster(mols, 'shape', returnDetails=True)
n_clusters = det['numClusters']
self.assertIsInstance(n_clusters,
np.int64,
msg="None valid number of clusters")
def test_01_cluster_maccs(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
mols = lib._mols
cl, det = cluster(mols, 'maccs', returnDetails=True)
ref_ncluster = MACCS_N_CLUSTER
ref_populations = MACCS_POPULATION_CLUSTER
ncluster = det['numClusters']
population = det['populations'].tolist()
self.assertEqual(
ncluster,
ref_ncluster,
msg="The number of cluster identified are not as expected")
self.assertEqual(
population,
ref_populations,
msg="The population fo the cluster are not the expected one")
def test_05_removeMols(self):
sdffile = os.path.join(self.dataDir, 'fda_drugs_light.sdf')
lib = SmallMolLib(sdffile)
mols_ids = SDF_IDS_DELETE
ref_mols_name = SDF_MOLNAME_DELETE
mols_name = [ s.ligname for s in lib.getMols(mols_ids) ]
self.assertListEqual(mols_name, ref_mols_name, msg="The molecules at the given indexes do not match with the"
"expected")
lib.removeMols(mols_ids)
mols_name_now = [ s.ligname for s in lib.getMols(mols_ids) ]
self.assertFalse(mols_name_now == mols_name, msg="The molecules seem to not be deleted correctly")