def test_nn_orders(self):
strat = OpenBabelNN()
acetylene = strat.get_nn_info(self.acetylene, 0)
self.assertEqual(acetylene[0]["weight"], 3)
self.assertEqual(acetylene[1]["weight"], 1)
# Currently, benzene bonds register either as double or single,
# not aromatic
# Instead of searching for aromatic bonds, we check that bonds are
# detected in the same way from both sides
self.assertEqual(strat.get_nn_info(self.benzene, 0)[0]["weight"],
strat.get_nn_info(self.benzene, 1)[0]["weight"])
def test_nn_orders(self):
strat = OpenBabelNN()
acetylene = strat.get_nn_info(self.acetylene, 0)
self.assertEqual(acetylene[0]["weight"], 3)
self.assertEqual(acetylene[1]["weight"], 1)
# Currently, benzene bonds register either as double or single,
# not aromatic
# Instead of searching for aromatic bonds, we check that bonds are
# detected in the same way from both sides
self.assertEqual(strat.get_nn_info(self.benzene, 0)[0]["weight"],
strat.get_nn_info(self.benzene, 1)[0]["weight"])
def test_nn_length(self):
strat = OpenBabelNN(order=False)
benzene_bonds = strat.get_nn_info(self.benzene, 0)
c_bonds = [b for b in benzene_bonds if str(b["site"].specie) == "C"]
h_bonds = [b for b in benzene_bonds if str(b["site"].specie) == "H"]
self.assertAlmostEqual(c_bonds[0]["weight"], 1.41, 2)
self.assertAlmostEqual(h_bonds[0]["weight"], 1.02, 2)
self.assertAlmostEqual(
strat.get_nn_info(self.acetylene, 0)[0]["weight"], 1.19, 2)
def test_nn_length(self):
strat = OpenBabelNN(order=False)
benzene_bonds = strat.get_nn_info(self.benzene, 0)
c_bonds = [b for b in benzene_bonds if str(b["site"].specie) == "C"]
h_bonds = [b for b in benzene_bonds if str(b["site"].specie) == "H"]
self.assertAlmostEqual(c_bonds[0]["weight"], 1.41, 2)
self.assertAlmostEqual(h_bonds[0]["weight"], 1.02, 2)
self.assertAlmostEqual(strat.get_nn_info(self.acetylene, 0)[0]["weight"],
1.19,
2)
def get_basic_functional_groups(self, func_groups=None):
"""
Identify functional groups that cannot be identified by the Ertl method
of get_special_carbon and get_heteroatoms, such as benzene rings, methyl
groups, and ethyl groups.
TODO: Think of other functional groups that are important enough to be
added (ex: do we need ethyl, butyl, propyl?)
:param func_groups: List of strs representing the functional groups of
interest. Default to None, meaning that all of the functional groups
defined in this function will be sought.
:return: list of sets of ints, representing groups of connected atoms
"""
strat = OpenBabelNN()
hydrogens = {
n
for n in self.molgraph.graph.nodes if str(self.species[n]) == "H"
}
carbons = [
n for n in self.molgraph.graph.nodes if str(self.species[n]) == "C"
]
if func_groups is None:
func_groups = ["methyl", "phenyl"]
results = []
if "methyl" in func_groups:
for node in carbons:
neighbors = strat.get_nn_info(self.molecule, node)
hs = {
n["site_index"]
for n in neighbors if n["site_index"] in hydrogens
}
# Methyl group is CH3, but this will also catch methane
if len(hs) >= 3:
hs.add(node)
results.append(hs)
if "phenyl" in func_groups:
rings_indices = [
set(sum(ring, ())) for ring in self.molgraph.find_rings()
]
possible_phenyl = [r for r in rings_indices if len(r) == 6]
for ring in possible_phenyl:
# Phenyl group should have only one (0 for benzene) member whose
# neighbors are not two carbons and one hydrogen
num_deviants = 0
for node in ring:
neighbors = strat.get_nn_info(self.molecule, node)
neighbor_spec = sorted([
str(self.species[n["site_index"]]) for n in neighbors
])
if neighbor_spec != ["C", "C", "H"]:
num_deviants += 1
if num_deviants <= 1:
for node in ring:
ring_group = copy.deepcopy(ring)
neighbors = self.molgraph.graph[node]
# Add hydrogens to the functional group
for neighbor in neighbors.keys():
if neighbor in hydrogens:
ring_group.add(neighbor)
results.append(ring_group)
return results
def get_basic_functional_groups(self, func_groups=None):
"""
Identify functional groups that cannot be identified by the Ertl method
of get_special_carbon and get_heteroatoms, such as benzene rings, methyl
groups, and ethyl groups.
TODO: Think of other functional groups that are important enough to be
added (ex: do we need ethyl, butyl, propyl?)
:param func_groups: List of strs representing the functional groups of
interest. Default to None, meaning that all of the functional groups
defined in this function will be sought.
:return: list of sets of ints, representing groups of connected atoms
"""
strat = OpenBabelNN()
hydrogens = {n for n in self.molgraph.graph.nodes if
str(self.species[n]) == "H"}
carbons = [n for n in self.molgraph.graph.nodes if
str(self.species[n]) == "C"]
if func_groups is None:
func_groups = ["methyl", "phenyl"]
results = []
if "methyl" in func_groups:
for node in carbons:
neighbors = strat.get_nn_info(self.molecule, node)
hs = {n["site_index"] for n in neighbors if n["site_index"] in hydrogens}
# Methyl group is CH3, but this will also catch methane
if len(hs) >= 3:
hs.add(node)
results.append(hs)
if "phenyl" in func_groups:
rings_indices = [set(sum(ring, ())) for ring in
self.molgraph.find_rings()]
possible_phenyl = [r for r in rings_indices if len(r) == 6]
for ring in possible_phenyl:
# Phenyl group should have only one (0 for benzene) member whose
# neighbors are not two carbons and one hydrogen
num_deviants = 0
for node in ring:
neighbors = strat.get_nn_info(self.molecule, node)
neighbor_spec = sorted([str(self.species[n["site_index"]])
for n in neighbors])
if neighbor_spec != ["C", "C", "H"]:
num_deviants += 1
if num_deviants <= 1:
for node in ring:
neighbors = self.molgraph.graph[node]
# Add hydrogens to the functional group
for neighbor in neighbors.keys():
if neighbor in hydrogens:
ring.add(neighbor)
results.append(ring)
return results
