def test_PatternHolder(self):
fname = os.path.join(os.environ["RDBASE"], "Data", "NCI", "first_5K.smi")
suppl = Chem.SmilesMolSupplier(fname, delimiter="\t", titleLine=False)
mols1 = rdSubstructLibrary.CachedTrustedSmilesMolHolder()
fps1 = rdSubstructLibrary.PatternHolder(2048)
ssslib1 = rdSubstructLibrary.SubstructLibrary(mols1, fps1)
mols2 = rdSubstructLibrary.CachedTrustedSmilesMolHolder()
fps2 = rdSubstructLibrary.PatternHolder()
ssslib2 = rdSubstructLibrary.SubstructLibrary(mols2, fps2)
RDLogger.DisableLog('rdApp.error')
for i in range(0, 1000, 10):
try:
mol = suppl[i]
except Exception:
continue
if (not mol):
continue
mols1.AddSmiles(Chem.MolToSmiles(mol))
fps1.AddFingerprint(fps1.MakeFingerprint(mol))
ssslib2.AddMol(mol)
RDLogger.EnableLog('rdApp.error')
query = Chem.MolFromSmarts("N")
self.assertIsNotNone(query)
matches1 = sorted(ssslib1.GetMatches(query))
matches2 = sorted(ssslib2.GetMatches(query))
self.assertEqual(len(matches1), len(matches2))
self.assertTrue(all([m1 == matches2[i] for i, m1 in enumerate(matches1)]))
def test_addpatterns(self):
pdb_ligands = [
"CCS(=O)(=O)c1ccc(OC)c(Nc2ncc(-c3cccc(-c4ccccn4)c3)o2)c1",
"COc1ccc(S(=O)(=O)NCC2CC2)cc1Nc1ncc(-c2cccc(-c3cccnc3)c2)o1",
"COc1ccc(-c2oc3ncnc(N)c3c2-c2ccc(NC(=O)Nc3cc(C(F)(F)F)ccc3F)cc2)cc1",
"COC(=O)Nc1nc2ccc(Oc3ccc(NC(=O)Nc4cc(C(F)(F)F)ccc4F)cc3)cc2[nH]1",
"COc1cc(Nc2ncnc(-c3cccnc3Nc3ccccc3)n2)cc(OC)c1OC",
"O=C(Nc1ccc(Oc2ccccc2)cc1)c1cccnc1NCc1ccncc1",
"O=C(Nc1ccc(Oc2ccccc2)cc1)c1cccnc1NCc1ccncc1",
"CNC(=O)c1cc(Oc2ccc3[nH]c(Nc4ccc(Cl)c(C(F)(F)F)c4)nc3c2)ccn1",
"CNC(=O)c1cc(Oc2ccc3oc(Nc4ccc(Cl)c(OCC5CCC[NH+]5C)c4)nc3c2)ccn1",
"CNC(=O)c1cc(Oc2ccc3oc(Nc4ccc(Cl)c(OCC5CCC[NH+]5C)c4)nc3c2)ccn1",
"COc1cc2nccc(Oc3ccc4c(c3)OCCN4C(=O)Nc3ccc(Cl)cc3)c2cc1OC",
"CNC(=O)c1c(C)oc2cc(Oc3cc[nH+]c4cc(OCCN5CCOCC5)ccc34)ccc12",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)Nc5ccc(Cl)cc5)cccc4c3)c2cc1OC",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)Nc5ccc(Cl)cc5)cccc4c3)c2cc1OC",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)NC5CC5)cccc4c3)c2cc1OC",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)NC5CC5)cccc4c3)c2cc1OC",
"Cc1ccc(C(=O)Nc2cc(CCC[NH+](C)C)cc(C(F)(F)F)c2)cc1Nc1ncccc1-c1ccncn1",
"COc1cc(Nc2nccc(Nc3ccc4c(C)n[nH]c4c3)n2)cc(OC)c1OC",
"COc1cc(Nc2nccc(N(C)c3ccc4c(C)n[nH]c4c3)n2)cc(OC)c1OC",
"Cc1ccn(-c2ccc3c(c2)NCC3(C)C)c(=O)c1-c1ccc2nc(N)ncc2c1",
"Cc1ccn(-c2ccc3c(c2)NCC3(C)C)c(=O)c1-c1ccc2nc(N)ncc2c1",
"Cc1ccc(C(=O)NCCC2CCCC2)cc1C(=O)Nc1ccc(N)nc1",
"Cc1ccc(C(=O)NCCC2CCCC2)cc1C(=O)Nc1ccc(N)nc1",
"Cc1ccn(-c2cccc(C(F)(F)F)c2)c(=O)c1-c1ccc2nc(N)ncc2c1",
"Cc1ccn(-c2cccc(C(F)(F)F)c2)c(=O)c1-c1ccc2nc(N)ncc2c1",
"O=C(Nc1cncnc1)c1c(Cl)ccc2c(Nc3cccc(C(F)(F)F)c3)noc12",
"O=C(Nc1cncnc1)c1c(Cl)ccc2c(Nc3cccc(C(F)(F)F)c3)noc12",
"CC1(C)CNc2cc(NC(=O)c3cccnc3NCc3ccncc3)ccc21",
"CC1(C)CNc2cc(NC(=O)c3cccnc3NCc3ccncc3)ccc21"
]
for patterns in [
rdSubstructLibrary.PatternHolder(),
rdSubstructLibrary.TautomerPatternHolder()
]:
mols = [Chem.MolFromSmiles(smi) for smi in pdb_ligands]
holder = rdSubstructLibrary.CachedMolHolder()
slib_with_patterns = rdSubstructLibrary.SubstructLibrary(
holder, patterns)
for mol in mols:
slib_with_patterns.AddMol(mol)
for nthreads in [1, 2, 0]:
slib_without_patterns = rdSubstructLibrary.SubstructLibrary(
holder, None)
rdSubstructLibrary.AddPatterns(slib_without_patterns, nthreads)
# check for seg fault
# were the fingerprints really created
slib_without_patterns.GetFpHolder().GetFingerprint(0)
for mol in mols:
l1 = slib_with_patterns.CountMatches(mol)
l2 = slib_without_patterns.CountMatches(mol)
self.assertTrue(l1)
self.assertEqual(l1, l2)
def test_basic_addpatterns(self):
# add mols
pdb_ligands = [
"CCS(=O)(=O)c1ccc(OC)c(Nc2ncc(-c3cccc(-c4ccccn4)c3)o2)c1",
"COc1ccc(S(=O)(=O)NCC2CC2)cc1Nc1ncc(-c2cccc(-c3cccnc3)c2)o1",
"COc1ccc(-c2oc3ncnc(N)c3c2-c2ccc(NC(=O)Nc3cc(C(F)(F)F)ccc3F)cc2)cc1",
"COC(=O)Nc1nc2ccc(Oc3ccc(NC(=O)Nc4cc(C(F)(F)F)ccc4F)cc3)cc2[nH]1",
"COc1cc(Nc2ncnc(-c3cccnc3Nc3ccccc3)n2)cc(OC)c1OC",
"O=C(Nc1ccc(Oc2ccccc2)cc1)c1cccnc1NCc1ccncc1",
"O=C(Nc1ccc(Oc2ccccc2)cc1)c1cccnc1NCc1ccncc1",
"CNC(=O)c1cc(Oc2ccc3[nH]c(Nc4ccc(Cl)c(C(F)(F)F)c4)nc3c2)ccn1",
"CNC(=O)c1cc(Oc2ccc3oc(Nc4ccc(Cl)c(OCC5CCC[NH+]5C)c4)nc3c2)ccn1",
"CNC(=O)c1cc(Oc2ccc3oc(Nc4ccc(Cl)c(OCC5CCC[NH+]5C)c4)nc3c2)ccn1",
"COc1cc2nccc(Oc3ccc4c(c3)OCCN4C(=O)Nc3ccc(Cl)cc3)c2cc1OC",
"CNC(=O)c1c(C)oc2cc(Oc3cc[nH+]c4cc(OCCN5CCOCC5)ccc34)ccc12",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)Nc5ccc(Cl)cc5)cccc4c3)c2cc1OC",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)Nc5ccc(Cl)cc5)cccc4c3)c2cc1OC",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)NC5CC5)cccc4c3)c2cc1OC",
"COc1cc2[nH+]ccc(Oc3ccc4c(C(=O)NC5CC5)cccc4c3)c2cc1OC",
"Cc1ccc(C(=O)Nc2cc(CCC[NH+](C)C)cc(C(F)(F)F)c2)cc1Nc1ncccc1-c1ccncn1",
"COc1cc(Nc2nccc(Nc3ccc4c(C)n[nH]c4c3)n2)cc(OC)c1OC",
"COc1cc(Nc2nccc(N(C)c3ccc4c(C)n[nH]c4c3)n2)cc(OC)c1OC",
"Cc1ccn(-c2ccc3c(c2)NCC3(C)C)c(=O)c1-c1ccc2nc(N)ncc2c1",
"Cc1ccn(-c2ccc3c(c2)NCC3(C)C)c(=O)c1-c1ccc2nc(N)ncc2c1",
"Cc1ccc(C(=O)NCCC2CCCC2)cc1C(=O)Nc1ccc(N)nc1",
"Cc1ccc(C(=O)NCCC2CCCC2)cc1C(=O)Nc1ccc(N)nc1",
"Cc1ccn(-c2cccc(C(F)(F)F)c2)c(=O)c1-c1ccc2nc(N)ncc2c1",
"Cc1ccn(-c2cccc(C(F)(F)F)c2)c(=O)c1-c1ccc2nc(N)ncc2c1",
"O=C(Nc1cncnc1)c1c(Cl)ccc2c(Nc3cccc(C(F)(F)F)c3)noc12",
"O=C(Nc1cncnc1)c1c(Cl)ccc2c(Nc3cccc(C(F)(F)F)c3)noc12",
"CC1(C)CNc2cc(NC(=O)c3cccnc3NCc3ccncc3)ccc21",
"CC1(C)CNc2cc(NC(=O)c3cccnc3NCc3ccncc3)ccc21"
]
for holder in [
rdSubstructLibrary.CachedSmilesMolHolder(),
rdSubstructLibrary.CachedTrustedSmilesMolHolder()
]:
for smi in pdb_ligands:
holder.AddSmiles(smi)
for patttern in [
None,
rdSubstructLibrary.PatternHolder(),
rdSubstructLibrary.TautomerPatternHolder()
]:
lib = rdSubstructLibrary.SubstructLibrary(holder)
rdSubstructLibrary.AddPatterns(lib, numThreads=-1)
self.assertEqual(len(lib.GetMolHolder()),
len(lib.GetFpHolder()))
for smi in pdb_ligands:
self.assertTrue(lib.CountMatches(Chem.MolFromSmiles(smi)))
def main(directory: str, chebml_version: str):
"""Download the ChEBML data."""
os.makedirs(directory, exist_ok=True)
bradley_path = os.path.join(directory, 'jm020472j_s2.xls')
if not os.path.exists(bradley_path):
try:
wget.download(bradley_url, out=directory)
except:
click.echo('There goes ACS stopping science')
chembl_url = (
f'ftp://ftp.ebi.ac.uk/pub/databases/chembl/ChEMBLdb/releases/'
f'chembl_{chebml_version}/chembl_{chebml_version}.sdf.gz')
sdf_path = os.path.join(directory, f'chembl_{chebml_version}.sdf.gz')
if not os.path.exists(sdf_path):
wget.download(chembl_url, out=directory)
sss_path = os.path.join(directory, f'chembl{chebml_version}_sssdata.pkl')
if not os.path.exists(sss_path):
click.echo(f'RDKit Version: {rdBase.rdkitVersion}')
data = []
with gzip.GzipFile(sdf_path) as gz:
suppl = Chem.ForwardSDMolSupplier(gz)
for mol in tqdm(suppl,
desc=f'Processing ChEBML {chebml_version}',
unit_scale=True):
if mol is None or mol.GetNumAtoms() > 50:
continue
fp = Chem.PatternFingerprint(mol)
smi = Chem.MolToSmiles(mol)
data.append((smi, fp))
click.echo(f'Outputting to {sss_path}')
with open(sss_path, 'wb') as file:
mols = rdSubstructLibrary.CachedTrustedSmilesMolHolder()
fps = rdSubstructLibrary.PatternHolder()
for smi, fp in data:
mols.AddSmiles(smi)
fps.AddFingerprint(fp)
library = rdSubstructLibrary.SubstructLibrary(mols, fps)
pickle.dump(library, file, protocol=pickle.HIGHEST_PROTOCOL)
click.echo('Done ;)')
def read_in_lib(input_smi):
t1 = time.time()
mols = rdSubstructLibrary.CachedTrustedSmilesMolHolder()
fps = rdSubstructLibrary.PatternHolder()
with open(input_smi, 'r') as inf:
ls = [x.split() for x in inf]
ls.pop(0)
with open(input_smi.replace(".smi", ".pkl"), 'rb') as pklf:
for l in tqdm.tqdm(ls):
smi = l[1]
mols.AddSmiles(smi)
fp = pickle.load(pklf)
fps.AddFingerprint(fp)
library = rdSubstructLibrary.SubstructLibrary(mols, fps)
t2 = time.time()
print("That took %.2f seconds. The library has %d molecules." %
(t2 - t1, len(library)))
return library
def __init__(self, proj, datapath, dbpath, chembldb, flimit=1e-3, MinClusterSize=20, clustering='UPGMA', calcDists=True, calcScores=False):
self.proj=proj
self.datapath=datapath
self.dbpath=dbpath
self.chembldb=chembldb
self.flimit=flimit
self.MinClusterSize=MinClusterSize
self.clustering=clustering
self.calcScores=calcScores
self.calcDists=calcDists
# load data
self.moldata_proj, self.distdata_proj=utilsDataPrep.PrepareData(self.proj,self.datapath,distMeasure='Tanimoto',FP='Morgan2', calcDists=self.calcDists)
if arthor is not None:
if not os.path.isdir(dbpath):
os.mkdir(dbpath)
# set up project database for arthor substructure matching
df=self.moldata_proj[['Structure','ID']]
df.to_csv('./arthor/{0}.smi'.format(self.proj), header=None, index=None, sep=' ')
os.system('smi2atdb -j 0 -t {0}{1}.smi {0}{1}.atdb'.format(self.dbpath,self.proj))
os.system('atdb2fp -j 0 {0}{1}.atdb'.format(self.dbpath,self.proj))
self.proj_db=arthor.SubDb('{0}{1}.atdb'.format(self.dbpath,self.proj))
else:
if type(dbpath)==rdSubstructLibrary.SubstructLibrary:
self.proj_db = dbpath
self.db_size = len(self.proj_db)
else:
if not os.path.exists(dbpath):
print("creating database")
mols = rdSubstructLibrary.CachedTrustedSmilesMolHolder()
fps = rdSubstructLibrary.PatternHolder()
for smi in self.moldata_proj['Structure']:
m = Chem.MolFromSmiles(smi)
mols.AddSmiles(Chem.MolToSmiles(m))
fps.AddFingerprint(Chem.PatternFingerprint(m))
self.proj_db = rdSubstructLibrary.SubstructLibrary(mols,fps)
self.db_size = len(mols)
pickle.dump(self.proj_db,open(dbpath,'wb+'))
else:
self.proj_db = pickle.load(open(dbpath,'rb'))
self.db_size = len(self.proj_db)