def celecoxib_similarity(smiles):
guac_benchmark = similarity(
smiles="CC1=CC=C(C=C1)C1=CC(=NN1C1=CC=C(C=C1)S(N)(=O)=O)C(F)(F)F",
name="Celecoxib",
fp_type="ECFP4",
threshold=1.0,
rediscovery=True,
)
return guac_benchmark.wrapped_objective.score(smiles)
def goal_directed_suite_v1() -> List[GoalDirectedBenchmark]:
max_logP = 6.35584
return [
isomers_c11h24(mean_function='arithmetic'),
isomers_c7h8n2o2(mean_function='arithmetic'),
isomers_c9h10n2o2pf2cl(mean_function='arithmetic', n_samples=100),
hard_cobimetinib(max_logP=max_logP),
hard_osimertinib(ArithmeticMeanScoringFunction),
hard_fexofenadine(ArithmeticMeanScoringFunction),
weird_physchem(),
# start pop benchmark
# e.g.
start_pop_ranolazine(),
# similarity Benchmarks
# explicit rediscovery
similarity(
smiles='CC1=CC=C(C=C1)C1=CC(=NN1C1=CC=C(C=C1)S(N)(=O)=O)C(F)(F)F',
name='Celecoxib',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True),
similarity(
smiles='Cc1c(C)c2OC(C)(COc3ccc(CC4SC(=O)NC4=O)cc3)CCc2c(C)c1O',
name='Troglitazone',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True),
similarity(
smiles='CN(C)S(=O)(=O)c1ccc2Sc3ccccc3C(=CCCN4CCN(C)CC4)c2c1',
name='Thiothixene',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True),
# generate similar stuff
similarity(smiles='Clc4cccc(N3CCN(CCCCOc2ccc1c(NC(=O)CC1)c2)CC3)c4Cl',
name='Aripiprazole',
fp_type='FCFP4',
threshold=0.75),
similarity(smiles='CC(C)(C)NCC(O)c1ccc(O)c(CO)c1',
name='Albuterol',
fp_type='FCFP4',
threshold=0.75),
similarity(
smiles=
'COc1ccc2[C@H]3CC[C@@]4(C)[C@@H](CC[C@@]4(O)C#C)[C@@H]3CCc2c1',
name='Mestranol',
fp_type='AP',
threshold=0.75),
logP_benchmark(target=-1.0),
logP_benchmark(target=8.0),
tpsa_benchmark(target=150.0),
cns_mpo(max_logP=max_logP),
qed_benchmark(),
median_camphor_menthol(ArithmeticMeanScoringFunction)
]
def goal_directed_suite_v2() -> List[GoalDirectedBenchmark]:
return [
# explicit rediscovery
similarity(
smiles='CC1=CC=C(C=C1)C1=CC(=NN1C1=CC=C(C=C1)S(N)(=O)=O)C(F)(F)F',
name='Celecoxib',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True),
similarity(
smiles='Cc1c(C)c2OC(C)(COc3ccc(CC4SC(=O)NC4=O)cc3)CCc2c(C)c1O',
name='Troglitazone',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True),
similarity(
smiles='CN(C)S(=O)(=O)c1ccc2Sc3ccccc3C(=CCCN4CCN(C)CC4)c2c1',
name='Thiothixene',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True),
# generate similar stuff
similarity(smiles='Clc4cccc(N3CCN(CCCCOc2ccc1c(NC(=O)CC1)c2)CC3)c4Cl',
name='Aripiprazole',
fp_type='ECFP4',
threshold=0.75),
similarity(smiles='CC(C)(C)NCC(O)c1ccc(O)c(CO)c1',
name='Albuterol',
fp_type='FCFP4',
threshold=0.75),
similarity(
smiles=
'COc1ccc2[C@H]3CC[C@@]4(C)[C@@H](CC[C@@]4(O)C#C)[C@@H]3CCc2c1',
name='Mestranol',
fp_type='AP',
threshold=0.75),
# isomers
isomers_c11h24(),
isomers_c9h10n2o2pf2cl(),
# median molecules
median_camphor_menthol(),
median_tadalafil_sildenafil(),
# all other MPOs
hard_osimertinib(),
hard_fexofenadine(),
ranolazine_mpo(),
perindopril_rings(),
amlodipine_rings(),
sitagliptin_replacement(),
zaleplon_with_other_formula(),
valsartan_smarts(),
decoration_hop(),
scaffold_hop(),
]
def get_guac_property_eval(self, task):
if task is GuacTask.CELECOXIB:
bench = standard_benchmarks.similarity(
smiles=
'CC1=CC=C(C=C1)C1=CC(=NN1C1=CC=C(C=C1)S(N)(=O)=O)C(F)(F)F',
name='Celecoxib',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True)
elif task is GuacTask.TROGLITAZONE:
bench = standard_benchmarks.similarity(
smiles='Cc1c(C)c2OC(C)(COc3ccc(CC4SC(=O)NC4=O)cc3)CCc2c(C)c1O',
name='Troglitazone',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True)
elif task is GuacTask.THIOTHIXENE:
bench = standard_benchmarks.similarity(
smiles='CN(C)S(=O)(=O)c1ccc2Sc3ccccc3C(=CCCN4CCN(C)CC4)c2c1',
name='Thiothixene',
fp_type='ECFP4',
threshold=1.0,
rediscovery=True)
elif task is GuacTask.ARIPIPRAZOLE:
bench = standard_benchmarks.similarity(
smiles='Clc4cccc(N3CCN(CCCCOc2ccc1c(NC(=O)CC1)c2)CC3)c4Cl',
name='Aripiprazole',
fp_type='ECFP4',
threshold=0.75)
elif task is GuacTask.ALBUTEROL:
bench = standard_benchmarks.similarity(
smiles='CC(C)(C)NCC(O)c1ccc(O)c(CO)c1',
name='Albuterol',
fp_type='FCFP4',
threshold=0.75)
elif task is GuacTask.MESTRANOL:
bench = standard_benchmarks.similarity(
smiles=
'COc1ccc2[C@H]3CC[C@@]4(C)[C@@H](CC[C@@]4(O)C#C)[C@@H]3CCc2c1',
name='Mestranol',
fp_type='AP',
threshold=0.75)
elif task is GuacTask.OSIMERTINIB:
bench = standard_benchmarks.hard_osimertinib()
elif task is GuacTask.RANOLAZINE:
bench = standard_benchmarks.ranolazine_mpo()
elif task is GuacTask.ZALEPLON:
bench = standard_benchmarks.zaleplon_with_other_formula()
elif task is GuacTask.VALSARTAN:
bench = standard_benchmarks.valsartan_smarts()
elif task is GuacTask.DECO:
bench = standard_benchmarks.decoration_hop()
elif task is GuacTask.SCAFFOLD:
bench = standard_benchmarks.scaffold_hop()
elif task is GuacTask.PERINDOPRIL:
bench = standard_benchmarks.perindopril_rings()
elif task is GuacTask.AMLODIPINE:
bench = standard_benchmarks.amlodipine_rings()
elif task is GuacTask.SITAGLIPTIN:
bench = standard_benchmarks.sitagliptin_replacement()
elif task is GuacTask.FEXOFENADINE:
bench = standard_benchmarks.hard_fexofenadine()
else:
raise NotImplementedError
smi2score = lambda smi: [bench.objective.score(smi)]
return PropertyEvaluator(smi2score)
