def test_nitrogen_charges(self):
"""
TestPDB: Verify that nitrogen groups are charged correctly.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# The labeling should pick up 2 charged nitrogen groups for two
# ammoniums.
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(
data_dir(), "ammonium_sulfate_hyd.pdbqt")
ammonium_sulfate_pdb.load_from_files(ammonium_sulfate_pdb_path,
ammonium_sulfate_pdbqt_path)
nitrogen_charges = ammonium_sulfate_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 2
assert nitrogen_charges[0].positive # Should be positive
assert nitrogen_charges[1].positive # Should be positive
# Test pyrrolidine (CH2)4NH. The nitrogen here should be sp3
# hybridized, so is likely to pick up an extra proton to its nitrogen
# at physiological pH.
pyrrolidine_pdb = PDB()
pyrrolidine_pdb_path = os.path.join(data_dir(), "pyrrolidine_hyd.pdb")
pyrrolidine_pdbqt_path = os.path.join(data_dir(),
"pyrrolidine_hyd.pdbqt")
pyrrolidine_pdb.load_from_files(pyrrolidine_pdb_path,
pyrrolidine_pdbqt_path)
nitrogen_charges = pyrrolidine_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 1
assert nitrogen_charges[0].positive # Should be positive
def featurize_dude(dude_dir, target, pickle_dir, num_jobs):
"""Featurize DUD-E docked poses and write features to pickle_out.
Parameters
----------
dude_dir: string
Path to DUD-E directory
target: string
Name of DUD-E target.
pickle_dir: string
Path to directory to output pickles
"""
target_dir = os.path.join(dude_dir, target)
actives_dir = os.path.join(target_dir, "actives")
decoys_dir = os.path.join(target_dir, "decoys")
actives = [a for a in os.listdir(actives_dir)]
decoys = [a for a in os.listdir(decoys_dir)]
receptor = os.path.join(target_dir, "receptor.pdb")
pickle_out = os.path.join(target_dir, "out.pkl.gz")
# Just for debugging purposes
actives = actives[:1]
num_per_job = int(math.ceil(len(actives) / float(num_jobs)))
print "Number per job: %d" % num_per_job
protein_pdb_path = "/home/rbharath/DUD-E/aa2ar/receptor_hyd.pdb"
protein_pdbqt_path = "/home/rbharath/DUD-E/aa2ar/receptor_hyd.pdbqt"
print "About to load protein from input files"
protein_pdb_obj = PDB()
protein_pdb_obj.load_from_files(protein_pdb_path, protein_pdbqt_path)
binana = Binana()
feature_len = binana.num_features()
feature_vectors = {}
for compound in actives:
compound_name = compound.split(".")[0]
compound_pdbqt = compound_name + "_hyd_out.pdbqt"
compound_pdbqt = os.path.join(actives_dir, compound_pdbqt)
# Convert the pdbqt to pdb
pdbqt_to_pdb(compound_pdbqt, actives_dir)
compound_pdb = compound_name + "_hyd_out.pdb"
compound_pdb = os.path.join(actives_dir, compound_pdb)
structures = MultiStructure()
structures.load_from_files(compound_pdb, compound_pdbqt)
vectors = []
for key in sorted(structures.molecules.keys()):
structure = structures.molecules[key]
print "type(structure)"
print type(structure)
vectors.append(
binana.compute_input_vector(structure, protein_pdb_obj))
feature_vectors[compound_name] = vectors
with gzip.open(pickle_out, "wb") as f:
pickle.dump(feature_vectors, f)
decoys = decoys[:1]
def test_metallic_charges(self):
"""
TestPDB: Verify that non-protein charges are assigned properly.
"""
# Test metallic ion charge.
magnesium_pdb = PDB()
magnesium_atom = Atom(element="MG",
coordinates=Point(coords=np.array([0,0,0])))
magnesium_pdb.add_new_non_protein_atom(magnesium_atom)
metallic_charges = magnesium_pdb.identify_metallic_charges()
assert len(metallic_charges) == 1
def test_metallic_charges(self):
"""
TestPDB: Verify that non-protein charges are assigned properly.
"""
# Test metallic ion charge.
magnesium_pdb = PDB()
magnesium_atom = Atom(element="MG",
coordinates=Point(coords=np.array([0, 0, 0])))
magnesium_pdb.add_new_non_protein_atom(magnesium_atom)
metallic_charges = magnesium_pdb.identify_metallic_charges()
assert len(metallic_charges) == 1
def test_assign_ligand_aromatics(self): """ TestPDB: Test that non-protein aromatic rings are assigned correctly. """ ### 3ao4 comes from PDBBind-CN and contains some cruft in the PDB file: ### atoms without residues labelled. This triggered some problems with ### non-protein aromatics complaining. # TODO(rbharath): Add a stub here. _3ao4_protein = PDB() _3ao4_protein_pdb = os.path.join(data_dir(), "3ao4_protein_hyd.pdb") _3ao4_protein_pdbqt = os.path.join(data_dir(), "3ao4_protein_hyd.pdbqt") _3ao4_protein.load_from_files(_3ao4_protein_pdb, _3ao4_protein_pdbqt)
def test_assign_ligand_aromatics(self):
"""
TestPDB: Test that non-protein aromatic rings are assigned correctly.
"""
### 3ao4 comes from PDBBind-CN and contains some cruft in the PDB file:
### atoms without residues labelled. This triggered some problems with
### non-protein aromatics complaining.
# TODO(rbharath): Add a stub here.
_3ao4_protein = PDB()
_3ao4_protein_pdb = os.path.join(data_dir(), "3ao4_protein_hyd.pdb")
_3ao4_protein_pdbqt = os.path.join(data_dir(),
"3ao4_protein_hyd.pdbqt")
_3ao4_protein.load_from_files(_3ao4_protein_pdb, _3ao4_protein_pdbqt)
def test_phosphorus_charges(self):
"""
TestPDB: Verify that Phosphorus groups are charged correctly.
"""
# CID82671 contains a phosphate between two aromatic groups.
phosphate_pdb = PDB()
phosphate_pdb_path = os.path.join(data_dir(), "82671_hyd.pdb")
phosphate_pdbqt_path = os.path.join(data_dir(), "82671_hyd.pdb")
phosphate_pdb.load_from_files(phosphate_pdb_path, phosphate_pdbqt_path)
phosphorus_charges = phosphate_pdb.identify_phosphorus_charges()
assert len(phosphorus_charges) == 1
assert not phosphorus_charges[
0].positive # Should be negatively charged.
def test_nitrogen_charges(self):
"""
TestPDB: Verify that nitrogen groups are charged correctly.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# The labeling should pick up 2 charged nitrogen groups for two
# ammoniums.
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdbqt")
ammonium_sulfate_pdb.load_from_files(
ammonium_sulfate_pdb_path, ammonium_sulfate_pdbqt_path)
nitrogen_charges = ammonium_sulfate_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 2
assert nitrogen_charges[0].positive # Should be positive
assert nitrogen_charges[1].positive # Should be positive
# Test pyrrolidine (CH2)4NH. The nitrogen here should be sp3
# hybridized, so is likely to pick up an extra proton to its nitrogen
# at physiological pH.
pyrrolidine_pdb = PDB()
pyrrolidine_pdb_path = os.path.join(data_dir(),
"pyrrolidine_hyd.pdb")
pyrrolidine_pdbqt_path = os.path.join(data_dir(),
"pyrrolidine_hyd.pdbqt")
pyrrolidine_pdb.load_from_files(pyrrolidine_pdb_path,
pyrrolidine_pdbqt_path)
nitrogen_charges = pyrrolidine_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 1
assert nitrogen_charges[0].positive # Should be positive
def test_load_bonds_from_pdb(self):
"""
TestPDB: Verifies that bonds can be loaded from PDB.
"""
pdb = PDB()
# Test that we can load CO2
carbon_atom = Atom(element="C")
oxygen_atom_1 = Atom(element="O")
oxygen_atom_2 = Atom(element="O")
pdb.add_new_atom(carbon_atom)
pdb.add_new_atom(oxygen_atom_1)
pdb.add_new_atom(oxygen_atom_2)
lines = [
"CONECT 1 2 3 "
"CONECT 2 "
"CONECT 3 "
]
with tempfile.NamedTemporaryFile() as temp:
temp.write("\n".join(lines))
temp.flush()
pdb.load_bonds_from_pdb(temp.name)
assert len(carbon_atom.indices_of_atoms_connecting) == 2
assert len(oxygen_atom_1.indices_of_atoms_connecting) == 0
assert len(oxygen_atom_2.indices_of_atoms_connecting) == 0
def test_sulfur_charges(self):
"""
TestPDB: Verify that sulfur groups are charged correctly.
"""
triflic_acid_pdb = PDB()
triflic_acid_pdb_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdb")
triflic_acid_pdbqt_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdbqt")
triflic_acid_pdb.load_from_files(triflic_acid_pdb_path,
triflic_acid_pdbqt_path)
sulfur_charges = (triflic_acid_pdb.identify_sulfur_charges())
assert len(sulfur_charges) == 1
assert not sulfur_charges[0].positive # Should be negatively charged.
def test_phosphorus_charges(self):
"""
TestPDB: Verify that Phosphorus groups are charged correctly.
"""
# CID82671 contains a phosphate between two aromatic groups.
phosphate_pdb = PDB()
phosphate_pdb_path = os.path.join(data_dir(),
"82671_hyd.pdb")
phosphate_pdbqt_path = os.path.join(data_dir(),
"82671_hyd.pdb")
phosphate_pdb.load_from_files(
phosphate_pdb_path, phosphate_pdbqt_path)
phosphorus_charges = phosphate_pdb.identify_phosphorus_charges()
assert len(phosphorus_charges) == 1
assert not phosphorus_charges[0].positive # Should be negatively charged.
def test_ligand_assign_aromatics(self):
"""
TestPDB: Verify that aromatic rings in ligands are identified.
"""
benzene_pdb = PDB()
benzene_pdb_path = os.path.join(data_dir(), "benzene_hyd.pdb")
benzene_pdbqt_path = os.path.join(data_dir(), "benzene_hyd.pdbqt")
benzene_pdb.load_from_files(benzene_pdb_path, benzene_pdbqt_path)
# A benzene should have exactly one aromatic ring.
print benzene_pdb.aromatic_rings
assert len(benzene_pdb.aromatic_rings) == 1
# The first 6 atoms in the benzene pdb form the aromatic ring.
assert (set(benzene_pdb.aromatic_rings[0].indices)
== set([1,2,3,4,5,6]))
def test_ligand_assign_aromatics(self):
"""
TestPDB: Verify that aromatic rings in ligands are identified.
"""
benzene_pdb = PDB()
benzene_pdb_path = os.path.join(data_dir(), "benzene_hyd.pdb")
benzene_pdbqt_path = os.path.join(data_dir(), "benzene_hyd.pdbqt")
benzene_pdb.load_from_files(benzene_pdb_path, benzene_pdbqt_path)
# A benzene should have exactly one aromatic ring.
print benzene_pdb.aromatic_rings
assert len(benzene_pdb.aromatic_rings) == 1
# The first 6 atoms in the benzene pdb form the aromatic ring.
assert (set(benzene_pdb.aromatic_rings[0].indices) == set(
[1, 2, 3, 4, 5, 6]))
def test_sulfur_charges(self):
"""
TestPDB: Verify that sulfur groups are charged correctly.
"""
triflic_acid_pdb = PDB()
triflic_acid_pdb_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdb")
triflic_acid_pdbqt_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdbqt")
triflic_acid_pdb.load_from_files(
triflic_acid_pdb_path,
triflic_acid_pdbqt_path)
sulfur_charges = (
triflic_acid_pdb.identify_sulfur_charges())
assert len(sulfur_charges) == 1
assert not sulfur_charges[0].positive # Should be negatively charged.
def test_load_bonds_from_pdb(self):
"""
TestPDB: Verifies that bonds can be loaded from PDB.
"""
pdb = PDB()
# Test that we can load CO2
carbon_atom = Atom(element="C")
oxygen_atom_1 = Atom(element="O")
oxygen_atom_2 = Atom(element="O")
pdb.add_new_atom(carbon_atom)
pdb.add_new_atom(oxygen_atom_1)
pdb.add_new_atom(oxygen_atom_2)
lines = [
"CONECT 1 2 3 "
"CONECT 2 "
"CONECT 3 "
]
with tempfile.NamedTemporaryFile() as temp:
temp.write("\n".join(lines))
temp.flush()
pdb.load_bonds_from_pdb(temp.name)
assert len(carbon_atom.indices_of_atoms_connecting) == 2
assert len(oxygen_atom_1.indices_of_atoms_connecting) == 0
assert len(oxygen_atom_2.indices_of_atoms_connecting) == 0
def compute_input_vector_from_files(self, ligand_pdb_filename,
receptor_pdb_filename, line_header):
"""Computes feature vector for ligand-receptor pair.
Parameters
----------
ligand_pdb_filename: string
path to ligand's pdb file.
receptor_pdb_filename: string
path to receptor pdb file.
line_header: string
line separator in PDB files
"""
# Load receptor and ligand from file.
receptor = PDB()
receptor.load_from_files(receptor_pdb_filename, line_header)
receptor.assign_secondary_structure()
ligand = PDB()
ligand.load_from_files(ligand_pdb_filename, line_header)
self.compute_input_vector(ligand, receptor)
def setUp(self):
"""
Instantiate a dummy PDB file.
"""
self.temp_dir = tempfile.mkdtemp()
self.pdb = PDB()
_, self.pdb_filename = tempfile.mkstemp(suffix=".pdb",
dir=self.temp_dir)
self.prgr_pdb = PDB()
prgr_pdb_path = os.path.join(data_dir(), "prgr_hyd.pdb")
prgr_pdbqt_path = os.path.join(data_dir(), "prgr_hyd.pdbqt")
self.prgr_pdb.load_from_files(prgr_pdb_path, prgr_pdbqt_path)
self._1r5y_protein = PDB()
_1r5y_protein_pdb = os.path.join(data_dir(), "1r5y_protein_hyd.pdb")
_1r5y_protein_pdbqt = os.path.join(data_dir(), "1r5y_protein_hyd.pdbqt")
self._1r5y_protein.load_from_files(_1r5y_protein_pdb, _1r5y_protein_pdbqt)
self.proteins = [("prgr", self.prgr_pdb), ("1r5y", self._1r5y_protein)]
def test_assign_non_protein_charges(self):
"""
TestPDB: Verify that charges are properly added to ligands.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# There should be 3 charged groups, two positive, one negative
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(
data_dir(), "ammonium_sulfate_hyd.pdbqt")
# Notice that load automatically identifies non-protein charges.
ammonium_sulfate_pdb.load_from_files(ammonium_sulfate_pdb_path,
ammonium_sulfate_pdbqt_path)
assert len(ammonium_sulfate_pdb.charges) == 3
num_pos, num_neg = 0, 0
for charge in ammonium_sulfate_pdb.charges:
if charge.positive:
num_pos += 1
else:
num_neg += 1
assert num_pos == 2
assert num_neg == 1
def test_assign_non_protein_charges(self):
"""
TestPDB: Verify that charges are properly added to ligands.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# There should be 3 charged groups, two positive, one negative
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdbqt")
# Notice that load automatically identifies non-protein charges.
ammonium_sulfate_pdb.load_from_files(
ammonium_sulfate_pdb_path, ammonium_sulfate_pdbqt_path)
assert len(ammonium_sulfate_pdb.charges) == 3
num_pos, num_neg = 0, 0
for charge in ammonium_sulfate_pdb.charges:
if charge.positive:
num_pos += 1
else:
num_neg += 1
assert num_pos == 2
assert num_neg == 1
def compute_input_vector_from_files(self, ligand_pdb_filename,
receptor_pdb_filename, line_header):
"""Computes feature vector for ligand-receptor pair.
Parameters
----------
ligand_pdb_filename: string
path to ligand's pdb file.
receptor_pdb_filename: string
path to receptor pdb file.
line_header: string
line separator in PDB files
"""
# Load receptor and ligand from file.
receptor = PDB()
receptor.load_from_files(receptor_pdb_filename, line_header)
receptor.assign_secondary_structure()
ligand = PDB()
ligand.load_from_files(ligand_pdb_filename, line_header)
self.compute_input_vector(ligand, receptor)
def _featurize_complex(self, mol_pdb, protein_pdb):
"""
Compute Binana fingerprint for complex.
"""
### OPEN TEMPDIR
tempdir = tempfile.mkdtemp()
mol_pdb_file = os.path.join(tempdir, "mol.pdb")
with open(mol_pdb_file, "w") as mol_f:
mol_f.writelines(mol_pdb)
protein_pdb_file = os.path.join(tempdir, "protein.pdb")
with open(protein_pdb_file, "w") as protein_f:
protein_f.writelines(protein_pdb)
mol_hyd_file = os.path.join(tempdir, "mol_hyd.pdb")
mol_pdbqt_file = os.path.join(tempdir, "mol_hyd.pdbqt")
hydrogenate_and_compute_partial_charges(
mol_pdb_file, "pdb", tempdir, mol_hyd_file, mol_pdbqt_file)
protein_hyd_file = os.path.join(tempdir, "protein_hyd.pdb")
protein_pdbqt_file = os.path.join(tempdir, "protein_hyd.pdbqt")
hydrogenate_and_compute_partial_charges(
protein_pdb_file, "pdb", tempdir, protein_hyd_file, protein_pdbqt_file)
mol_pdb_obj = PDB()
mol_pdb_obj.load_from_files(mol_pdb_file, mol_pdbqt_file)
protein_pdb_obj = PDB()
protein_pdb_obj.load_from_files(protein_pdb_file, protein_pdbqt_file)
features = self.binana.compute_input_vector(mol_pdb_obj, protein_pdb_obj)
### CLOSE TEMPDIR
shutil.rmtree(tempdir)
return features
def setUp(self):
"""
Instantiate a dummy PDB file.
"""
self.temp_dir = tempfile.mkdtemp()
self.pdb = PDB()
_, self.pdb_filename = tempfile.mkstemp(suffix=".pdb",
dir=self.temp_dir)
self.prgr_pdb = PDB()
prgr_pdb_path = os.path.join(data_dir(), "prgr_hyd.pdb")
prgr_pdbqt_path = os.path.join(data_dir(), "prgr_hyd.pdbqt")
self.prgr_pdb.load_from_files(prgr_pdb_path, prgr_pdbqt_path)
self._1r5y_protein = PDB()
_1r5y_protein_pdb = os.path.join(data_dir(), "1r5y_protein_hyd.pdb")
_1r5y_protein_pdbqt = os.path.join(data_dir(),
"1r5y_protein_hyd.pdbqt")
self._1r5y_protein.load_from_files(_1r5y_protein_pdb,
_1r5y_protein_pdbqt)
self.proteins = [("prgr", self.prgr_pdb), ("1r5y", self._1r5y_protein)]
def featurize_pdbbind(pdbbind_dir, pickle_out):
"""Featurize all entries in pdbbind_dir and write features to pickle_out
pdbbind_dir should be a dir, with K subdirs, one for each protein-ligand
complex to be featurized. The ligand and receptor should each have a pdb
and pdbqt file. The ligand files should end in '_ligand_hyd.${FILETYPE}'
while the receptor files should end in '_protein_hyd.${FILETYPE}'
pdbbind_dir: string
Path to pdbbind directory.
pickle_out: string
Path to write pickle output.
"""
assert os.path.isdir(pdbbind_dir)
# Instantiate copy of binana vector
binana = Binana()
feature_vectors = {}
# Extract the subdirectories in pdbbind_dir
subdirs = [d for d in os.listdir(pdbbind_dir) if
os.path.isdir(os.path.join(pdbbind_dir, d))]
# TODO(rbharath): ONLY FOR DEBUGGING!
subdirs = subdirs[900:]
num_atoms = len(Binana.atom_types)
# See features/tests/nnscore_test.py:TestBinana.testComputeInputVector
# for derivation.
feature_len = (3*num_atoms*(num_atoms+1)/2 + num_atoms + 12 + 6 + 3 + 6 +
3 + 6 + 3 + 1)
for count, d in enumerate(subdirs):
print "\nprocessing %d-th pdb %s" % (count, d)
subdir = os.path.join(pdbbind_dir, d)
print "About to extract ligand and protein input files"
ligand_pdb, ligand_pdbqt = None, None
protein_pdb, protein_pdbqt = None, None
for f in os.listdir(subdir):
if re.search("_ligand_hyd.pdb$", f):
ligand_pdb = f
elif re.search("_ligand_hyd.pdbqt$", f):
ligand_pdbqt = f
elif re.search("_protein_hyd.pdb$", f):
protein_pdb = f
elif re.search("_protein_hyd.pdbqt$", f):
protein_pdbqt = f
print "Extracted Input Files:"
print (ligand_pdb, ligand_pdbqt, protein_pdb, protein_pdbqt)
if (not ligand_pdb or not ligand_pdbqt or not protein_pdb or not
protein_pdbqt):
raise ValueError("Required files not present for %s" % d)
ligand_pdb_path = os.path.join(subdir, ligand_pdb)
ligand_pdbqt_path = os.path.join(subdir, ligand_pdbqt)
protein_pdb_path = os.path.join(subdir, protein_pdb)
protein_pdbqt_path = os.path.join(subdir, protein_pdbqt)
print "About to load ligand from input files"
ligand_pdb_obj = PDB()
ligand_pdb_obj.load_from_files(ligand_pdb_path, ligand_pdbqt_path)
print "About to load protein from input files"
protein_pdb_obj = PDB()
protein_pdb_obj.load_from_files(protein_pdb_path, protein_pdbqt_path)
print "About to generate feature vector."
vector = binana.compute_input_vector(ligand_pdb_obj,
protein_pdb_obj)
feature_vectors[d] = vector
if len(vector) != feature_len:
raise ValueError("Feature length incorrect on %s" % d)
print "Feature vector generated correctly."
with open(pickle_out, "wb") as f:
pickle.dump(feature_vectors, f)
def setUp(self):
"""
Instantiate local copy of Binana object.
"""
self.binana = Binana()
### 3zp9 comes from PDBBind-CN
_3zp9_protein = PDB()
_3zp9_protein_pdb = os.path.join(data_dir(), "3zp9_protein_hyd.pdb")
_3zp9_protein_pdbqt = os.path.join(data_dir(), "3zp9_protein_hyd.pdbqt")
_3zp9_protein.load_from_files(_3zp9_protein_pdb, _3zp9_protein_pdbqt)
# The ligand is also specified by pdbbind
_3zp9_ligand = PDB()
_3zp9_ligand_pdb = os.path.join(data_dir(), "3zp9_ligand_hyd.pdb")
_3zp9_ligand_pdbqt = os.path.join(data_dir(), "3zp9_ligand_hyd.pdbqt")
_3zp9_ligand.load_from_files(_3zp9_ligand_pdb, _3zp9_ligand_pdbqt)
### 3bwf comes from PDBBind-CN
_3bwf_protein = PDB()
_3bwf_protein_pdb = os.path.join(data_dir(), "3bwf_protein_hyd.pdb")
_3bwf_protein_pdbqt = os.path.join(data_dir(), "3bwf_protein_hyd.pdbqt")
_3bwf_protein.load_from_files(_3bwf_protein_pdb, _3bwf_protein_pdbqt)
# The ligand is also specified by pdbbind
_3bwf_ligand = PDB()
_3bwf_ligand_pdb = os.path.join(data_dir(), "3bwf_ligand_hyd.pdb")
_3bwf_ligand_pdbqt = os.path.join(data_dir(), "3bwf_ligand_hyd.pdbqt")
_3bwf_ligand.load_from_files(_3bwf_ligand_pdb, _3bwf_ligand_pdbqt)
self.test_cases = [("3bwf", _3bwf_protein, _3bwf_ligand),
("3zp9", _3zp9_protein, _3zp9_ligand)]
def setUp(self):
"""
Instantiate local copy of Binana object.
"""
self.binana = Binana()
# TODO(rbharath): This load sequence is getting unwieldy. Refactor.
#### PRGR is from the DUD-E collection
#prgr_receptor = PDB()
#prgr_pdb = os.path.join(data_dir(), "prgr_hyd.pdb")
#prgr_pdbqt = os.path.join(data_dir(), "prgr_hyd.pdbqt")
#prgr_receptor.load_from_files(prgr_pdb, prgr_pdbqt)
## This compound is CHEMBL1164248
#prgr_active = PDB()
#prgr_active_pdb = os.path.join(data_dir(), "prgr_active0_hyd.pdb")
#prgr_active_pdbqt = os.path.join(data_dir(), "prgr_active0_hyd.pdbqt")
#prgr_active.load_from_files(prgr_active_pdb, prgr_active_pdbqt)
#### c-Abl is taken from the Autodock Vina examples
#cabl_receptor = PDB()
#cabl_receptor_pdb = os.path.join(data_dir(), "c-Abl_hyd.pdb")
#cabl_receptor_pdbqt = os.path.join(data_dir(), "c-Abl_hyd.pdbqt")
#cabl_receptor.load_from_files(cabl_receptor_pdb,
# cabl_receptor_pdbqt)
## This compound is imatinib
#cabl_active = PDB()
#cabl_active_pdb = os.path.join(data_dir(), "imatinib_hyd.pdb")
#cabl_active_pdbqt = os.path.join(data_dir(), "imatinib_hyd.pdbqt")
#cabl_active.load_from_files(cabl_active_pdb,
# cabl_active_pdbqt)
#### 1zea comes from PDBBind-CN
## Python complains about variables starting with numbers, so put an
## underscore in front of everything.
#_1zea_protein = PDB()
#_1zea_protein_pdb = os.path.join(data_dir(), "1zea_protein_hyd.pdb")
#_1zea_protein_pdbqt = os.path.join(data_dir(), "1zea_protein_hyd.pdbqt")
#_1zea_protein.load_from_files(_1zea_protein_pdb, _1zea_protein_pdbqt)
## The ligand is also specified by pdbbind
#_1zea_ligand = PDB()
#_1zea_ligand_pdb = os.path.join(data_dir(), "1zea_ligand_hyd.pdb")
#_1zea_ligand_pdbqt = os.path.join(data_dir(), "1zea_ligand_hyd.pdbqt")
#_1zea_ligand.load_from_files(_1zea_ligand_pdb, _1zea_ligand_pdbqt)
#### 1r5y comes from PDBBind-CN
#_1r5y_protein = PDB()
#_1r5y_protein_pdb = os.path.join(data_dir(), "1r5y_protein_hyd.pdb")
#_1r5y_protein_pdbqt = os.path.join(data_dir(), "1r5y_protein_hyd.pdbqt")
#_1r5y_protein.load_from_files(_1r5y_protein_pdb, _1r5y_protein_pdbqt)
## The ligand is also specified by pdbbind
#_1r5y_ligand = PDB()
#_1r5y_ligand_pdb = os.path.join(data_dir(), "1r5y_ligand_hyd.pdb")
#_1r5y_ligand_pdbqt = os.path.join(data_dir(), "1r5y_ligand_hyd.pdbqt")
#_1r5y_ligand.load_from_files(_1r5y_ligand_pdb, _1r5y_ligand_pdbqt)
#### 3ao4 comes from PDBBind-CN
#_3ao4_protein = PDB()
#_3ao4_protein_pdb = os.path.join(data_dir(), "3ao4_protein_hyd.pdb")
#_3ao4_protein_pdbqt = os.path.join(data_dir(), "3ao4_protein_hyd.pdbqt")
#_3ao4_protein.load_from_files(_3ao4_protein_pdb, _3ao4_protein_pdbqt)
## The ligand is also specified by pdbbind
#_3ao4_ligand = PDB()
#_3ao4_ligand_pdb = os.path.join(data_dir(), "3ao4_ligand_hyd.pdb")
#_3ao4_ligand_pdbqt = os.path.join(data_dir(), "3ao4_ligand_hyd.pdbqt")
#_3ao4_ligand.load_from_files(_3ao4_ligand_pdb, _3ao4_ligand_pdbqt)
#### 2jdm comes from PDBBind-CN
#_2jdm_protein = PDB()
#_2jdm_protein_pdb = os.path.join(data_dir(), "2jdm_protein_hyd.pdb")
#_2jdm_protein_pdbqt = os.path.join(data_dir(), "2jdm_protein_hyd.pdbqt")
#_2jdm_protein.load_from_files(_2jdm_protein_pdb, _2jdm_protein_pdbqt)
## The ligand is also specified by pdbbind
#_2jdm_ligand = PDB()
#_2jdm_ligand_pdb = os.path.join(data_dir(), "2jdm_ligand_hyd.pdb")
#_2jdm_ligand_pdbqt = os.path.join(data_dir(), "2jdm_ligand_hyd.pdbqt")
#_2jdm_ligand.load_from_files(_2jdm_ligand_pdb, _2jdm_ligand_pdbqt)
#### 3g5k comes from PDBBind-CN
#_3g5k_protein = PDB()
#_3g5k_protein_pdb = os.path.join(data_dir(), "3g5k_protein_hyd.pdb")
#_3g5k_protein_pdbqt = os.path.join(data_dir(), "3g5k_protein_hyd.pdbqt")
#_3g5k_protein.load_from_files(_3g5k_protein_pdb, _3g5k_protein_pdbqt)
## The ligand is also specified by pdbbind
#_3g5k_ligand = PDB()
#_3g5k_ligand_pdb = os.path.join(data_dir(), "3g5k_ligand_hyd.pdb")
#_3g5k_ligand_pdbqt = os.path.join(data_dir(), "3g5k_ligand_hyd.pdbqt")
#_3g5k_ligand.load_from_files(_3g5k_ligand_pdb, _3g5k_ligand_pdbqt)
#### 3str comes from PDBBind-CN
#_3str_protein = PDB()
#_3str_protein_pdb = os.path.join(data_dir(), "3str_protein_hyd.pdb")
#_3str_protein_pdbqt = os.path.join(data_dir(), "3str_protein_hyd.pdbqt")
#_3str_protein.load_from_files(_3str_protein_pdb, _3str_protein_pdbqt)
## The ligand is also specified by pdbbind
#_3str_ligand = PDB()
#_3str_ligand_pdb = os.path.join(data_dir(), "3str_ligand_hyd.pdb")
#_3str_ligand_pdbqt = os.path.join(data_dir(), "3str_ligand_hyd.pdbqt")
#_3str_ligand.load_from_files(_3str_ligand_pdb, _3str_ligand_pdbqt)
#### 1nu3 comes from PDBBind-CN
#_1nu3_protein = PDB()
#_1nu3_protein_pdb = os.path.join(data_dir(), "1nu3_protein_hyd.pdb")
#_1nu3_protein_pdbqt = os.path.join(data_dir(), "1nu3_protein_hyd.pdbqt")
#_1nu3_protein.load_from_files(_1nu3_protein_pdb, _1nu3_protein_pdbqt)
## The ligand is also specified by pdbbind
#_1nu3_ligand = PDB()
#_1nu3_ligand_pdb = os.path.join(data_dir(), "1nu3_ligand_hyd.pdb")
#_1nu3_ligand_pdbqt = os.path.join(data_dir(), "1nu3_ligand_hyd.pdbqt")
#_1nu3_ligand.load_from_files(_1nu3_ligand_pdb, _1nu3_ligand_pdbqt)
#### 2rio comes from PDBBind-CN
#_2rio_protein = PDB()
#_2rio_protein_pdb = os.path.join(data_dir(), "2rio_protein_hyd.pdb")
#_2rio_protein_pdbqt = os.path.join(data_dir(), "2rio_protein_hyd.pdbqt")
#_2rio_protein.load_from_files(_2rio_protein_pdb, _2rio_protein_pdbqt)
## The ligand is also specified by pdbbind
#_2rio_ligand = PDB()
#_2rio_ligand_pdb = os.path.join(data_dir(), "2rio_ligand_hyd.pdb")
#_2rio_ligand_pdbqt = os.path.join(data_dir(), "2rio_ligand_hyd.pdbqt")
#_2rio_ligand.load_from_files(_2rio_ligand_pdb, _2rio_ligand_pdbqt)
#### 2y2h comes from PDBBind-CN
#_2y2h_protein = PDB()
#_2y2h_protein_pdb = os.path.join(data_dir(), "2y2h_protein_hyd.pdb")
#_2y2h_protein_pdbqt = os.path.join(data_dir(), "2y2h_protein_hyd.pdbqt")
#_2y2h_protein.load_from_files(_2y2h_protein_pdb, _2y2h_protein_pdbqt)
## The ligand is also specified by pdbbind
#_2y2h_ligand = PDB()
#_2y2h_ligand_pdb = os.path.join(data_dir(), "2y2h_ligand_hyd.pdb")
#_2y2h_ligand_pdbqt = os.path.join(data_dir(), "2y2h_ligand_hyd.pdbqt")
#_2y2h_ligand.load_from_files(_2y2h_ligand_pdb, _2y2h_ligand_pdbqt)
#### 1pi5 comes from PDBBind-CN
#_1pi5_protein = PDB()
#_1pi5_protein_pdb = os.path.join(data_dir(), "1pi5_protein_hyd.pdb")
#_1pi5_protein_pdbqt = os.path.join(data_dir(), "1pi5_protein_hyd.pdbqt")
#_1pi5_protein.load_from_files(_1pi5_protein_pdb, _1pi5_protein_pdbqt)
## The ligand is also specified by pdbbind
#_1pi5_ligand = PDB()
#_1pi5_ligand_pdb = os.path.join(data_dir(), "1pi5_ligand_hyd.pdb")
#_1pi5_ligand_pdbqt = os.path.join(data_dir(), "1pi5_ligand_hyd.pdbqt")
#_1pi5_ligand.load_from_files(_1pi5_ligand_pdb, _1pi5_ligand_pdbqt)
#### 3fxz comes from PDBBind-CN
#_3fxz_protein = PDB()
#_3fxz_protein_pdb = os.path.join(data_dir(), "3fxz_protein_hyd.pdb")
#_3fxz_protein_pdbqt = os.path.join(data_dir(), "3fxz_protein_hyd.pdbqt")
#_3fxz_protein.load_from_files(_3fxz_protein_pdb, _3fxz_protein_pdbqt)
## The ligand is also specified by pdbbind
#_3fxz_ligand = PDB()
#_3fxz_ligand_pdb = os.path.join(data_dir(), "3fxz_ligand_hyd.pdb")
#_3fxz_ligand_pdbqt = os.path.join(data_dir(), "3fxz_ligand_hyd.pdbqt")
#_3fxz_ligand.load_from_files(_3fxz_ligand_pdb, _3fxz_ligand_pdbqt)
#### 4i60 comes from PDBBind-CN
#_4i60_protein = PDB()
#_4i60_protein_pdb = os.path.join(data_dir(), "4i60_protein_hyd.pdb")
#_4i60_protein_pdbqt = os.path.join(data_dir(), "4i60_protein_hyd.pdbqt")
#_4i60_protein.load_from_files(_4i60_protein_pdb, _4i60_protein_pdbqt)
## The ligand is also specified by pdbbind
#_4i60_ligand = PDB()
#_4i60_ligand_pdb = os.path.join(data_dir(), "4i60_ligand_hyd.pdb")
#_4i60_ligand_pdbqt = os.path.join(data_dir(), "4i60_ligand_hyd.pdbqt")
#_4i60_ligand.load_from_files(_4i60_ligand_pdb, _4i60_ligand_pdbqt)
#### 1hyv comes from PDBBind-CN
#_1hyv_protein = PDB()
#_1hyv_protein_pdb = os.path.join(data_dir(), "1hyv_protein_hyd.pdb")
#_1hyv_protein_pdbqt = os.path.join(data_dir(), "1hyv_protein_hyd.pdbqt")
#_1hyv_protein.load_from_files(_1hyv_protein_pdb, _1hyv_protein_pdbqt)
## The ligand is also specified by pdbbind
#_1hyv_ligand = PDB()
#_1hyv_ligand_pdb = os.path.join(data_dir(), "1hyv_ligand_hyd.pdb")
#_1hyv_ligand_pdbqt = os.path.join(data_dir(), "1hyv_ligand_hyd.pdbqt")
#_1hyv_ligand.load_from_files(_1hyv_ligand_pdb, _1hyv_ligand_pdbqt)
#### 3m1j comes from PDBBind-CN
#_3m1j_protein = PDB()
#_3m1j_protein_pdb = os.path.join(data_dir(), "3m1j_protein_hyd.pdb")
#_3m1j_protein_pdbqt = os.path.join(data_dir(), "3m1j_protein_hyd.pdbqt")
#_3m1j_protein.load_from_files(_3m1j_protein_pdb, _3m1j_protein_pdbqt)
## The ligand is also specified by pdbbind
#_3m1j_ligand = PDB()
#_3m1j_ligand_pdb = os.path.join(data_dir(), "3m1j_ligand_hyd.pdb")
#_3m1j_ligand_pdbqt = os.path.join(data_dir(), "3m1j_ligand_hyd.pdbqt")
#_3m1j_ligand.load_from_files(_3m1j_ligand_pdb, _3m1j_ligand_pdbqt)
#### 1y3g comes from PDBBind-CN
#_1y3g_protein = PDB()
#_1y3g_protein_pdb = os.path.join(data_dir(), "1y3g_protein_hyd.pdb")
#_1y3g_protein_pdbqt = os.path.join(data_dir(), "1y3g_protein_hyd.pdbqt")
#_1y3g_protein.load_from_files(_1y3g_protein_pdb, _1y3g_protein_pdbqt)
## The ligand is also specified by pdbbind
#_1y3g_ligand = PDB()
#_1y3g_ligand_pdb = os.path.join(data_dir(), "1y3g_ligand_hyd.pdb")
#_1y3g_ligand_pdbqt = os.path.join(data_dir(), "1y3g_ligand_hyd.pdbqt")
#_1y3g_ligand.load_from_files(_1y3g_ligand_pdb, _1y3g_ligand_pdbqt)
#### 6rsa comes from PDBBind-CN
#_6rsa_protein = PDB()
#_6rsa_protein_pdb = os.path.join(data_dir(), "6rsa_protein_hyd.pdb")
#_6rsa_protein_pdbqt = os.path.join(data_dir(), "6rsa_protein_hyd.pdbqt")
#_6rsa_protein.load_from_files(_6rsa_protein_pdb, _6rsa_protein_pdbqt)
## The ligand is also specified by pdbbind
#_6rsa_ligand = PDB()
#_6rsa_ligand_pdb = os.path.join(data_dir(), "6rsa_ligand_hyd.pdb")
#_6rsa_ligand_pdbqt = os.path.join(data_dir(), "6rsa_ligand_hyd.pdbqt")
#_6rsa_ligand.load_from_files(_6rsa_ligand_pdb, _6rsa_ligand_pdbqt)
#### 1lvk comes from PDBBind-CN
#_1lvk_protein = PDB()
#_1lvk_protein_pdb = os.path.join(data_dir(), "1lvk_protein_hyd.pdb")
#_1lvk_protein_pdbqt = os.path.join(data_dir(), "1lvk_protein_hyd.pdbqt")
#_1lvk_protein.load_from_files(_1lvk_protein_pdb, _1lvk_protein_pdbqt)
## The ligand is also specified by pdbbind
#_1lvk_ligand = PDB()
#_1lvk_ligand_pdb = os.path.join(data_dir(), "1lvk_ligand_hyd.pdb")
#_1lvk_ligand_pdbqt = os.path.join(data_dir(), "1lvk_ligand_hyd.pdbqt")
#_1lvk_ligand.load_from_files(_1lvk_ligand_pdb, _1lvk_ligand_pdbqt)
#### 3rj7 comes from PDBBind-CN
#_3rj7_protein = PDB()
#_3rj7_protein_pdb = os.path.join(data_dir(), "3rj7_protein_hyd.pdb")
#_3rj7_protein_pdbqt = os.path.join(data_dir(), "3rj7_protein_hyd.pdbqt")
#_3rj7_protein.load_from_files(_3rj7_protein_pdb, _3rj7_protein_pdbqt)
## The ligand is also specified by pdbbind
#_3rj7_ligand = PDB()
#_3rj7_ligand_pdb = os.path.join(data_dir(), "3rj7_ligand_hyd.pdb")
#_3rj7_ligand_pdbqt = os.path.join(data_dir(), "3rj7_ligand_hyd.pdbqt")
#_3rj7_ligand.load_from_files(_3rj7_ligand_pdb, _3rj7_ligand_pdbqt)
### 3zp9 comes from PDBBind-CN
_3zp9_protein = PDB()
_3zp9_protein_pdb = os.path.join(data_dir(), "3zp9_protein_hyd.pdb")
_3zp9_protein_pdbqt = os.path.join(data_dir(), "3zp9_protein_hyd.pdbqt")
_3zp9_protein.load_from_files(_3zp9_protein_pdb, _3zp9_protein_pdbqt)
# The ligand is also specified by pdbbind
_3zp9_ligand = PDB()
_3zp9_ligand_pdb = os.path.join(data_dir(), "3zp9_ligand_hyd.pdb")
_3zp9_ligand_pdbqt = os.path.join(data_dir(), "3zp9_ligand_hyd.pdbqt")
_3zp9_ligand.load_from_files(_3zp9_ligand_pdb, _3zp9_ligand_pdbqt)
### 3bwf comes from PDBBind-CN
_3bwf_protein = PDB()
_3bwf_protein_pdb = os.path.join(data_dir(), "3bwf_protein_hyd.pdb")
_3bwf_protein_pdbqt = os.path.join(data_dir(), "3bwf_protein_hyd.pdbqt")
_3bwf_protein.load_from_files(_3bwf_protein_pdb, _3bwf_protein_pdbqt)
# The ligand is also specified by pdbbind
_3bwf_ligand = PDB()
_3bwf_ligand_pdb = os.path.join(data_dir(), "3bwf_ligand_hyd.pdb")
_3bwf_ligand_pdbqt = os.path.join(data_dir(), "3bwf_ligand_hyd.pdbqt")
_3bwf_ligand.load_from_files(_3bwf_ligand_pdb, _3bwf_ligand_pdbqt)
#self.test_cases = [("prgr", prgr_receptor, prgr_active),
# ("cabl", cabl_receptor, cabl_active),
# ("1zea", _1zea_protein, _1zea_ligand),
# ("1r5y", _1r5y_protein, _1r5y_ligand),
# ("3ao4", _3ao4_protein, _3ao4_ligand),
# ("2jdm", _2jdm_protein, _2jdm_ligand),
# ("3g5k", _3g5k_protein, _3g5k_ligand),
# ("3str", _3str_protein, _3str_ligand),
# ("2rio", _2rio_protein, _2rio_ligand)]
self.test_cases = [("3bwf", _3bwf_protein, _3bwf_ligand)]
def setUp(self):
"""
Instantiate local copy of Binana object.
"""
self.binana = Binana()
# TODO(rbharath): This load sequence is getting unwieldy. Refactor.
### PRGR is from the DUD-E collection
prgr_receptor = PDB()
prgr_pdb = os.path.join(data_dir(), "prgr_hyd.pdb")
prgr_pdbqt = os.path.join(data_dir(), "prgr_hyd.pdbqt")
prgr_receptor.load_from_files(prgr_pdb, prgr_pdbqt)
# This compound is CHEMBL1164248
prgr_active = PDB()
prgr_active_pdb = os.path.join(data_dir(), "prgr_active0_hyd.pdb")
prgr_active_pdbqt = os.path.join(data_dir(), "prgr_active0_hyd.pdbqt")
prgr_active.load_from_files(prgr_active_pdb, prgr_active_pdbqt)
### c-Abl is taken from the Autodock Vina examples
cabl_receptor = PDB()
cabl_receptor_pdb = os.path.join(data_dir(), "c-Abl_hyd.pdb")
cabl_receptor_pdbqt = os.path.join(data_dir(), "c-Abl_hyd.pdbqt")
cabl_receptor.load_from_files(cabl_receptor_pdb,
cabl_receptor_pdbqt)
# This compound is imatinib
cabl_active = PDB()
cabl_active_pdb = os.path.join(data_dir(), "imatinib_hyd.pdb")
cabl_active_pdbqt = os.path.join(data_dir(), "imatinib_hyd.pdbqt")
cabl_active.load_from_files(cabl_active_pdb,
cabl_active_pdbqt)
### 1zea comes from PDBBind-CN
# Python complains about variables starting with numbers, so put an
# underscore in front of everything.
_1zea_protein = PDB()
_1zea_protein_pdb = os.path.join(data_dir(), "1zea_protein_hyd.pdb")
_1zea_protein_pdbqt = os.path.join(data_dir(), "1zea_protein_hyd.pdbqt")
_1zea_protein.load_from_files(_1zea_protein_pdb, _1zea_protein_pdbqt)
# The ligand is also specified by pdbbind
_1zea_ligand = PDB()
_1zea_ligand_pdb = os.path.join(data_dir(), "1zea_ligand_hyd.pdb")
_1zea_ligand_pdbqt = os.path.join(data_dir(), "1zea_ligand_hyd.pdbqt")
_1zea_ligand.load_from_files(_1zea_ligand_pdb, _1zea_ligand_pdbqt)
### 1r5y comes from PDBBind-CN
_1r5y_protein = PDB()
_1r5y_protein_pdb = os.path.join(data_dir(), "1r5y_protein_hyd.pdb")
_1r5y_protein_pdbqt = os.path.join(data_dir(), "1r5y_protein_hyd.pdbqt")
_1r5y_protein.load_from_files(_1r5y_protein_pdb, _1r5y_protein_pdbqt)
# The ligand is also specified by pdbbind
_1r5y_ligand = PDB()
_1r5y_ligand_pdb = os.path.join(data_dir(), "1r5y_ligand_hyd.pdb")
_1r5y_ligand_pdbqt = os.path.join(data_dir(), "1r5y_ligand_hyd.pdbqt")
_1r5y_ligand.load_from_files(_1r5y_ligand_pdb, _1r5y_ligand_pdbqt)
### 3ao4 comes from PDBBind-CN
_3ao4_protein = PDB()
_3ao4_protein_pdb = os.path.join(data_dir(), "3ao4_protein_hyd.pdb")
_3ao4_protein_pdbqt = os.path.join(data_dir(), "3ao4_protein_hyd.pdbqt")
_3ao4_protein.load_from_files(_3ao4_protein_pdb, _3ao4_protein_pdbqt)
# The ligand is also specified by pdbbind
_3ao4_ligand = PDB()
_3ao4_ligand_pdb = os.path.join(data_dir(), "3ao4_ligand_hyd.pdb")
_3ao4_ligand_pdbqt = os.path.join(data_dir(), "3ao4_ligand_hyd.pdbqt")
_3ao4_ligand.load_from_files(_3ao4_ligand_pdb, _3ao4_ligand_pdbqt)
### 2jdm comes from PDBBind-CN
_2jdm_protein = PDB()
_2jdm_protein_pdb = os.path.join(data_dir(), "2jdm_protein_hyd.pdb")
_2jdm_protein_pdbqt = os.path.join(data_dir(), "2jdm_protein_hyd.pdbqt")
_2jdm_protein.load_from_files(_2jdm_protein_pdb, _2jdm_protein_pdbqt)
# The ligand is also specified by pdbbind
_2jdm_ligand = PDB()
_2jdm_ligand_pdb = os.path.join(data_dir(), "2jdm_ligand_hyd.pdb")
_2jdm_ligand_pdbqt = os.path.join(data_dir(), "2jdm_ligand_hyd.pdbqt")
_2jdm_ligand.load_from_files(_2jdm_ligand_pdb, _2jdm_ligand_pdbqt)
### 3g5k comes from PDBBind-CN
_3g5k_protein = PDB()
_3g5k_protein_pdb = os.path.join(data_dir(), "3g5k_protein_hyd.pdb")
_3g5k_protein_pdbqt = os.path.join(data_dir(), "3g5k_protein_hyd.pdbqt")
_3g5k_protein.load_from_files(_3g5k_protein_pdb, _3g5k_protein_pdbqt)
# The ligand is also specified by pdbbind
_3g5k_ligand = PDB()
_3g5k_ligand_pdb = os.path.join(data_dir(), "3g5k_ligand_hyd.pdb")
_3g5k_ligand_pdbqt = os.path.join(data_dir(), "3g5k_ligand_hyd.pdbqt")
_3g5k_ligand.load_from_files(_3g5k_ligand_pdb, _3g5k_ligand_pdbqt)
### 3str comes from PDBBind-CN
_3str_protein = PDB()
_3str_protein_pdb = os.path.join(data_dir(), "3str_protein_hyd.pdb")
_3str_protein_pdbqt = os.path.join(data_dir(), "3str_protein_hyd.pdbqt")
_3str_protein.load_from_files(_3str_protein_pdb, _3str_protein_pdbqt)
# The ligand is also specified by pdbbind
_3str_ligand = PDB()
_3str_ligand_pdb = os.path.join(data_dir(), "3str_ligand_hyd.pdb")
_3str_ligand_pdbqt = os.path.join(data_dir(), "3str_ligand_hyd.pdbqt")
_3str_ligand.load_from_files(_3str_ligand_pdb, _3str_ligand_pdbqt)
### 1nu3 comes from PDBBind-CN
_1nu3_protein = PDB()
_1nu3_protein_pdb = os.path.join(data_dir(), "1nu3_protein_hyd.pdb")
_1nu3_protein_pdbqt = os.path.join(data_dir(), "1nu3_protein_hyd.pdbqt")
_1nu3_protein.load_from_files(_1nu3_protein_pdb, _1nu3_protein_pdbqt)
# The ligand is also specified by pdbbind
_1nu3_ligand = PDB()
_1nu3_ligand_pdb = os.path.join(data_dir(), "1nu3_ligand_hyd.pdb")
_1nu3_ligand_pdbqt = os.path.join(data_dir(), "1nu3_ligand_hyd.pdbqt")
_1nu3_ligand.load_from_files(_1nu3_ligand_pdb, _1nu3_ligand_pdbqt)
### 2rio comes from PDBBind-CN
_2rio_protein = PDB()
_2rio_protein_pdb = os.path.join(data_dir(), "2rio_protein_hyd.pdb")
_2rio_protein_pdbqt = os.path.join(data_dir(), "2rio_protein_hyd.pdbqt")
_2rio_protein.load_from_files(_2rio_protein_pdb, _2rio_protein_pdbqt)
# The ligand is also specified by pdbbind
_2rio_ligand = PDB()
_2rio_ligand_pdb = os.path.join(data_dir(), "2rio_ligand_hyd.pdb")
_2rio_ligand_pdbqt = os.path.join(data_dir(), "2rio_ligand_hyd.pdbqt")
_2rio_ligand.load_from_files(_2rio_ligand_pdb, _2rio_ligand_pdbqt)
self.test_cases = [("prgr", prgr_receptor, prgr_active),
("cabl", cabl_receptor, cabl_active),
("1zea", _1zea_protein, _1zea_ligand),
("1r5y", _1r5y_protein, _1r5y_ligand),
("3ao4", _3ao4_protein, _3ao4_ligand),
("2jdm", _2jdm_protein, _2jdm_ligand),
("3g5k", _3g5k_protein, _3g5k_ligand),
("3str", _3str_protein, _3str_ligand),
("2rio", _2rio_protein, _2rio_ligand)]
def setUp(self):
"""
Instantiate local copy of Binana object.
"""
self.binana = Binana()
# TODO(rbharath): This load sequence is getting unwieldy. Refactor.
### PRGR is from the DUD-E collection
prgr_receptor = PDB()
prgr_pdb = os.path.join(data_dir(), "prgr_hyd.pdb")
prgr_pdbqt = os.path.join(data_dir(), "prgr_hyd.pdbqt")
prgr_receptor.load_from_files(prgr_pdb, prgr_pdbqt)
# This compound is CHEMBL1164248
prgr_active = PDB()
prgr_active_pdb = os.path.join(data_dir(), "prgr_active0_hyd.pdb")
prgr_active_pdbqt = os.path.join(data_dir(), "prgr_active0_hyd.pdbqt")
prgr_active.load_from_files(prgr_active_pdb, prgr_active_pdbqt)
### c-Abl is taken from the Autodock Vina examples
cabl_receptor = PDB()
cabl_receptor_pdb = os.path.join(data_dir(), "c-Abl_hyd.pdb")
cabl_receptor_pdbqt = os.path.join(data_dir(), "c-Abl_hyd.pdbqt")
cabl_receptor.load_from_files(cabl_receptor_pdb, cabl_receptor_pdbqt)
# This compound is imatinib
cabl_active = PDB()
cabl_active_pdb = os.path.join(data_dir(), "imatinib_hyd.pdb")
cabl_active_pdbqt = os.path.join(data_dir(), "imatinib_hyd.pdbqt")
cabl_active.load_from_files(cabl_active_pdb, cabl_active_pdbqt)
### 1zea comes from PDBBind-CN
# Python complains about variables starting with numbers, so put an
# underscore in front of everything.
_1zea_protein = PDB()
_1zea_protein_pdb = os.path.join(data_dir(), "1zea_protein_hyd.pdb")
_1zea_protein_pdbqt = os.path.join(data_dir(),
"1zea_protein_hyd.pdbqt")
_1zea_protein.load_from_files(_1zea_protein_pdb, _1zea_protein_pdbqt)
# The ligand is also specified by pdbbind
_1zea_ligand = PDB()
_1zea_ligand_pdb = os.path.join(data_dir(), "1zea_ligand_hyd.pdb")
_1zea_ligand_pdbqt = os.path.join(data_dir(), "1zea_ligand_hyd.pdbqt")
_1zea_ligand.load_from_files(_1zea_ligand_pdb, _1zea_ligand_pdbqt)
### 1r5y comes from PDBBind-CN
_1r5y_protein = PDB()
_1r5y_protein_pdb = os.path.join(data_dir(), "1r5y_protein_hyd.pdb")
_1r5y_protein_pdbqt = os.path.join(data_dir(),
"1r5y_protein_hyd.pdbqt")
_1r5y_protein.load_from_files(_1r5y_protein_pdb, _1r5y_protein_pdbqt)
# The ligand is also specified by pdbbind
_1r5y_ligand = PDB()
_1r5y_ligand_pdb = os.path.join(data_dir(), "1r5y_ligand_hyd.pdb")
_1r5y_ligand_pdbqt = os.path.join(data_dir(), "1r5y_ligand_hyd.pdbqt")
_1r5y_ligand.load_from_files(_1r5y_ligand_pdb, _1r5y_ligand_pdbqt)
### 3ao4 comes from PDBBind-CN
_3ao4_protein = PDB()
_3ao4_protein_pdb = os.path.join(data_dir(), "3ao4_protein_hyd.pdb")
_3ao4_protein_pdbqt = os.path.join(data_dir(),
"3ao4_protein_hyd.pdbqt")
_3ao4_protein.load_from_files(_3ao4_protein_pdb, _3ao4_protein_pdbqt)
# The ligand is also specified by pdbbind
_3ao4_ligand = PDB()
_3ao4_ligand_pdb = os.path.join(data_dir(), "3ao4_ligand_hyd.pdb")
_3ao4_ligand_pdbqt = os.path.join(data_dir(), "3ao4_ligand_hyd.pdbqt")
_3ao4_ligand.load_from_files(_3ao4_ligand_pdb, _3ao4_ligand_pdbqt)
### 2jdm comes from PDBBind-CN
_2jdm_protein = PDB()
_2jdm_protein_pdb = os.path.join(data_dir(), "2jdm_protein_hyd.pdb")
_2jdm_protein_pdbqt = os.path.join(data_dir(),
"2jdm_protein_hyd.pdbqt")
_2jdm_protein.load_from_files(_2jdm_protein_pdb, _2jdm_protein_pdbqt)
# The ligand is also specified by pdbbind
_2jdm_ligand = PDB()
_2jdm_ligand_pdb = os.path.join(data_dir(), "2jdm_ligand_hyd.pdb")
_2jdm_ligand_pdbqt = os.path.join(data_dir(), "2jdm_ligand_hyd.pdbqt")
_2jdm_ligand.load_from_files(_2jdm_ligand_pdb, _2jdm_ligand_pdbqt)
### 3g5k comes from PDBBind-CN
_3g5k_protein = PDB()
_3g5k_protein_pdb = os.path.join(data_dir(), "3g5k_protein_hyd.pdb")
_3g5k_protein_pdbqt = os.path.join(data_dir(),
"3g5k_protein_hyd.pdbqt")
_3g5k_protein.load_from_files(_3g5k_protein_pdb, _3g5k_protein_pdbqt)
# The ligand is also specified by pdbbind
_3g5k_ligand = PDB()
_3g5k_ligand_pdb = os.path.join(data_dir(), "3g5k_ligand_hyd.pdb")
_3g5k_ligand_pdbqt = os.path.join(data_dir(), "3g5k_ligand_hyd.pdbqt")
_3g5k_ligand.load_from_files(_3g5k_ligand_pdb, _3g5k_ligand_pdbqt)
### 3str comes from PDBBind-CN
_3str_protein = PDB()
_3str_protein_pdb = os.path.join(data_dir(), "3str_protein_hyd.pdb")
_3str_protein_pdbqt = os.path.join(data_dir(),
"3str_protein_hyd.pdbqt")
_3str_protein.load_from_files(_3str_protein_pdb, _3str_protein_pdbqt)
# The ligand is also specified by pdbbind
_3str_ligand = PDB()
_3str_ligand_pdb = os.path.join(data_dir(), "3str_ligand_hyd.pdb")
_3str_ligand_pdbqt = os.path.join(data_dir(), "3str_ligand_hyd.pdbqt")
_3str_ligand.load_from_files(_3str_ligand_pdb, _3str_ligand_pdbqt)
### 1nu3 comes from PDBBind-CN
_1nu3_protein = PDB()
_1nu3_protein_pdb = os.path.join(data_dir(), "1nu3_protein_hyd.pdb")
_1nu3_protein_pdbqt = os.path.join(data_dir(),
"1nu3_protein_hyd.pdbqt")
_1nu3_protein.load_from_files(_1nu3_protein_pdb, _1nu3_protein_pdbqt)
# The ligand is also specified by pdbbind
_1nu3_ligand = PDB()
_1nu3_ligand_pdb = os.path.join(data_dir(), "1nu3_ligand_hyd.pdb")
_1nu3_ligand_pdbqt = os.path.join(data_dir(), "1nu3_ligand_hyd.pdbqt")
_1nu3_ligand.load_from_files(_1nu3_ligand_pdb, _1nu3_ligand_pdbqt)
### 2rio comes from PDBBind-CN
_2rio_protein = PDB()
_2rio_protein_pdb = os.path.join(data_dir(), "2rio_protein_hyd.pdb")
_2rio_protein_pdbqt = os.path.join(data_dir(),
"2rio_protein_hyd.pdbqt")
_2rio_protein.load_from_files(_2rio_protein_pdb, _2rio_protein_pdbqt)
# The ligand is also specified by pdbbind
_2rio_ligand = PDB()
_2rio_ligand_pdb = os.path.join(data_dir(), "2rio_ligand_hyd.pdb")
_2rio_ligand_pdbqt = os.path.join(data_dir(), "2rio_ligand_hyd.pdbqt")
_2rio_ligand.load_from_files(_2rio_ligand_pdb, _2rio_ligand_pdbqt)
self.test_cases = [("prgr", prgr_receptor, prgr_active),
("cabl", cabl_receptor, cabl_active),
("1zea", _1zea_protein, _1zea_ligand),
("1r5y", _1r5y_protein, _1r5y_ligand),
("3ao4", _3ao4_protein, _3ao4_ligand),
("2jdm", _2jdm_protein, _2jdm_ligand),
("3g5k", _3g5k_protein, _3g5k_ligand),
("3str", _3str_protein, _3str_ligand),
("2rio", _2rio_protein, _2rio_ligand)]
def featurize_dude(dude_dir, target, pickle_dir, num_jobs):
"""Featurize DUD-E docked poses and write features to pickle_out.
Parameters
----------
dude_dir: string
Path to DUD-E directory
target: string
Name of DUD-E target.
pickle_dir: string
Path to directory to output pickles
"""
target_dir = os.path.join(dude_dir, target)
actives_dir = os.path.join(target_dir, "actives")
decoys_dir = os.path.join(target_dir, "decoys")
actives = [a for a in os.listdir(actives_dir)]
decoys = [a for a in os.listdir(decoys_dir)]
receptor = os.path.join(target_dir, "receptor.pdb")
pickle_out = os.path.join(target_dir, "out.pkl.gz")
# Just for debugging purposes
actives = actives[:1]
num_per_job = int(math.ceil(len(actives)/float(num_jobs)))
print "Number per job: %d" % num_per_job
protein_pdb_path = "/home/rbharath/DUD-E/aa2ar/receptor_hyd.pdb"
protein_pdbqt_path = "/home/rbharath/DUD-E/aa2ar/receptor_hyd.pdbqt"
print "About to load protein from input files"
protein_pdb_obj = PDB()
protein_pdb_obj.load_from_files(protein_pdb_path, protein_pdbqt_path)
binana = Binana()
feature_len = binana.num_features()
feature_vectors = {}
for compound in actives:
compound_name = compound.split(".")[0]
compound_pdbqt = compound_name + "_hyd_out.pdbqt"
compound_pdbqt = os.path.join(actives_dir, compound_pdbqt)
# Convert the pdbqt to pdb
pdbqt_to_pdb(compound_pdbqt, actives_dir)
compound_pdb = compound_name + "_hyd_out.pdb"
compound_pdb = os.path.join(actives_dir, compound_pdb)
structures = MultiStructure()
structures.load_from_files(compound_pdb, compound_pdbqt)
vectors = []
for key in sorted(structures.molecules.keys()):
structure = structures.molecules[key]
print "type(structure)"
print type(structure)
vectors.append(binana.compute_input_vector(structure,
protein_pdb_obj))
feature_vectors[compound_name] = vectors
with gzip.open(pickle_out, "wb") as f:
pickle.dump(feature_vectors, f)
decoys = decoys[:1]
def test_carbon_charges(self):
"""
TestPDB: Verify that carbon groups are charged correctly.
"""
# Guanidine is positively charged at physiological pH
guanidine_pdb = PDB()
guanidine_pdb_path = os.path.join(data_dir(),
"guanidine_hyd.pdb")
guanidine_pdbqt_path = os.path.join(data_dir(),
"guanidine_hyd.pdbqt")
guanidine_pdb.load_from_files(
guanidine_pdb_path, guanidine_pdbqt_path)
carbon_charges = guanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# sulfaguanidine contains a guanidine group that is likely to be
# positively protonated at physiological pH
sulfaguanidine_pdb = PDB()
sulfaguanidine_pdb_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdb")
sulfaguanidine_pdbqt_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdbqt")
sulfaguanidine_pdb.load_from_files(
sulfaguanidine_pdb_path, sulfaguanidine_pdbqt_path)
carbon_charges = sulfaguanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# Formic acid is a carboxylic acid, which should be negatively charged.
formic_acid_pdb = PDB()
formic_acid_pdb_path = os.path.join(data_dir(),
"formic_acid_hyd.pdb")
formic_acid_pdbqt_path = os.path.join(data_dir(),
"formic_acid_hyd.pdbqt")
formic_acid_pdb.load_from_files(
formic_acid_pdb_path, formic_acid_pdbqt_path)
carbon_charges = formic_acid_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert not carbon_charges[0].positive # Should be negatively charged.
def setUp(self):
"""
Instantiate local copy of Binana object.
"""
self.binana = Binana()
### 3zp9 comes from PDBBind-CN
_3zp9_protein = PDB()
_3zp9_protein_pdb = os.path.join(data_dir(), "3zp9_protein_hyd.pdb")
_3zp9_protein_pdbqt = os.path.join(data_dir(),
"3zp9_protein_hyd.pdbqt")
_3zp9_protein.load_from_files(_3zp9_protein_pdb, _3zp9_protein_pdbqt)
# The ligand is also specified by pdbbind
_3zp9_ligand = PDB()
_3zp9_ligand_pdb = os.path.join(data_dir(), "3zp9_ligand_hyd.pdb")
_3zp9_ligand_pdbqt = os.path.join(data_dir(), "3zp9_ligand_hyd.pdbqt")
_3zp9_ligand.load_from_files(_3zp9_ligand_pdb, _3zp9_ligand_pdbqt)
### 3bwf comes from PDBBind-CN
_3bwf_protein = PDB()
_3bwf_protein_pdb = os.path.join(data_dir(), "3bwf_protein_hyd.pdb")
_3bwf_protein_pdbqt = os.path.join(data_dir(),
"3bwf_protein_hyd.pdbqt")
_3bwf_protein.load_from_files(_3bwf_protein_pdb, _3bwf_protein_pdbqt)
# The ligand is also specified by pdbbind
_3bwf_ligand = PDB()
_3bwf_ligand_pdb = os.path.join(data_dir(), "3bwf_ligand_hyd.pdb")
_3bwf_ligand_pdbqt = os.path.join(data_dir(), "3bwf_ligand_hyd.pdbqt")
_3bwf_ligand.load_from_files(_3bwf_ligand_pdb, _3bwf_ligand_pdbqt)
self.test_cases = [("3bwf", _3bwf_protein, _3bwf_ligand),
("3zp9", _3zp9_protein, _3zp9_ligand)]
def test_carbon_charges(self):
"""
TestPDB: Verify that carbon groups are charged correctly.
"""
# Guanidine is positively charged at physiological pH
guanidine_pdb = PDB()
guanidine_pdb_path = os.path.join(data_dir(), "guanidine_hyd.pdb")
guanidine_pdbqt_path = os.path.join(data_dir(), "guanidine_hyd.pdbqt")
guanidine_pdb.load_from_files(guanidine_pdb_path, guanidine_pdbqt_path)
carbon_charges = guanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# sulfaguanidine contains a guanidine group that is likely to be
# positively protonated at physiological pH
sulfaguanidine_pdb = PDB()
sulfaguanidine_pdb_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdb")
sulfaguanidine_pdbqt_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdbqt")
sulfaguanidine_pdb.load_from_files(sulfaguanidine_pdb_path,
sulfaguanidine_pdbqt_path)
carbon_charges = sulfaguanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# Formic acid is a carboxylic acid, which should be negatively charged.
formic_acid_pdb = PDB()
formic_acid_pdb_path = os.path.join(data_dir(), "formic_acid_hyd.pdb")
formic_acid_pdbqt_path = os.path.join(data_dir(),
"formic_acid_hyd.pdbqt")
formic_acid_pdb.load_from_files(formic_acid_pdb_path,
formic_acid_pdbqt_path)
carbon_charges = formic_acid_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert not carbon_charges[0].positive # Should be negatively charged.
def featurize_pdbbind(pdbbind_dir, pickle_out):
"""Featurize all entries in pdbbind_dir and write features to pickle_out
pdbbind_dir should be a dir, with K subdirs, one for each protein-ligand
complex to be featurized. The ligand and receptor should each have a pdb
and pdbqt file. The ligand files should end in '_ligand_hyd.${FILETYPE}'
while the receptor files should end in '_protein_hyd.${FILETYPE}'
pdbbind_dir: string
Path to pdbbind directory.
pickle_out: string
Path to write pickle output.
"""
assert os.path.isdir(pdbbind_dir)
# Instantiate copy of binana vector
binana = Binana()
feature_vectors = {}
# Extract the subdirectories in pdbbind_dir
subdirs = [
d for d in os.listdir(pdbbind_dir)
if os.path.isdir(os.path.join(pdbbind_dir, d))
]
# TODO(rbharath): ONLY FOR DEBUGGING!
subdirs = subdirs[900:]
num_atoms = len(Binana.atom_types)
# See features/tests/nnscore_test.py:TestBinana.testComputeInputVector
# for derivation.
feature_len = (3 * num_atoms * (num_atoms + 1) / 2 + num_atoms + 12 + 6 +
3 + 6 + 3 + 6 + 3 + 1)
for count, d in enumerate(subdirs):
print "\nprocessing %d-th pdb %s" % (count, d)
subdir = os.path.join(pdbbind_dir, d)
print "About to extract ligand and protein input files"
ligand_pdb, ligand_pdbqt = None, None
protein_pdb, protein_pdbqt = None, None
for f in os.listdir(subdir):
if re.search("_ligand_hyd.pdb$", f):
ligand_pdb = f
elif re.search("_ligand_hyd.pdbqt$", f):
ligand_pdbqt = f
elif re.search("_protein_hyd.pdb$", f):
protein_pdb = f
elif re.search("_protein_hyd.pdbqt$", f):
protein_pdbqt = f
print "Extracted Input Files:"
print(ligand_pdb, ligand_pdbqt, protein_pdb, protein_pdbqt)
if (not ligand_pdb or not ligand_pdbqt or not protein_pdb
or not protein_pdbqt):
raise ValueError("Required files not present for %s" % d)
ligand_pdb_path = os.path.join(subdir, ligand_pdb)
ligand_pdbqt_path = os.path.join(subdir, ligand_pdbqt)
protein_pdb_path = os.path.join(subdir, protein_pdb)
protein_pdbqt_path = os.path.join(subdir, protein_pdbqt)
print "About to load ligand from input files"
ligand_pdb_obj = PDB()
ligand_pdb_obj.load_from_files(ligand_pdb_path, ligand_pdbqt_path)
print "About to load protein from input files"
protein_pdb_obj = PDB()
protein_pdb_obj.load_from_files(protein_pdb_path, protein_pdbqt_path)
print "About to generate feature vector."
vector = binana.compute_input_vector(ligand_pdb_obj, protein_pdb_obj)
feature_vectors[d] = vector
if len(vector) != feature_len:
raise ValueError("Feature length incorrect on %s" % d)
print "Feature vector generated correctly."
with open(pickle_out, "wb") as f:
pickle.dump(feature_vectors, f)
class TestPDB(unittest.TestCase):
""""
Test PDB class.
"""
def setUp(self):
"""
Instantiate a dummy PDB file.
"""
self.temp_dir = tempfile.mkdtemp()
self.pdb = PDB()
_, self.pdb_filename = tempfile.mkstemp(suffix=".pdb",
dir=self.temp_dir)
self.prgr_pdb = PDB()
prgr_pdb_path = os.path.join(data_dir(), "prgr_hyd.pdb")
prgr_pdbqt_path = os.path.join(data_dir(), "prgr_hyd.pdbqt")
self.prgr_pdb.load_from_files(prgr_pdb_path, prgr_pdbqt_path)
self._1r5y_protein = PDB()
_1r5y_protein_pdb = os.path.join(data_dir(), "1r5y_protein_hyd.pdb")
_1r5y_protein_pdbqt = os.path.join(data_dir(),
"1r5y_protein_hyd.pdbqt")
self._1r5y_protein.load_from_files(_1r5y_protein_pdb,
_1r5y_protein_pdbqt)
self.proteins = [("prgr", self.prgr_pdb), ("1r5y", self._1r5y_protein)]
def tearDown(self):
"""
Delete temporary directory.
"""
shutil.rmtree(self.temp_dir)
def test_add_new_atom(self):
"""
TestPDB: Verifies that new atoms can be added.
"""
# Verify that no atoms are present when we start.
assert len(self.pdb.all_atoms.keys()) == 0
empty_atom = Atom()
self.pdb.add_new_atom(empty_atom)
# Verify that we now have one atom
assert len(self.pdb.all_atoms.keys()) == 1
def test_get_residues(self):
"""
TestPDB: Tests that all residues in PDB are identified.
"""
residues = self.prgr_pdb.get_residues()
# prgr.pdb has 280 unique residues
assert len(residues.keys()) == 280
prgr_residues = [
"LEU", "ILE", "ASN", "LEU", "LEU", "MET", "SER", "ILE", "GLU",
"PRO", "ASP", "VAL", "ILE", "TYR", "ALA", "GLY", "HIS", "ASP",
"THR", "SER", "SER", "SER", "LEU", "LEU", "THR", "SER", "LEU",
"ASN", "GLN", "LEU", "GLY", "GLU", "ARG", "GLN", "LEU", "LEU",
"SER", "VAL", "VAL", "LYS", "TRP", "SER", "LYS", "SER", "LEU",
"PRO", "GLY", "PHE", "ARG", "LEU", "HIS", "ILE", "ASP", "ASP",
"GLN", "ILE", "THR", "LEU", "ILE", "GLN", "TYR", "SER", "TRP",
"MET", "SER", "LEU", "MET", "VAL", "PHE", "GLY", "LEU", "GLY",
"TRP", "ARG", "SER", "TYR", "LYS", "HIS", "VAL", "SER", "GLY",
"GLN", "MET", "LEU", "TYR", "PHE", "ALA", "PRO", "ASP", "LEU",
"ILE", "LEU", "ASN", "GLU", "GLN", "ARG", "MET", "LYS", "GLU",
"PHE", "TYR", "SER", "LEU", "CYS", "LEU", "THR", "MET", "TRP",
"GLN", "ILE", "PRO", "GLN", "GLU", "PHE", "VAL", "LYS", "LEU",
"GLN", "VAL", "SER", "GLN", "GLU", "GLU", "PHE", "LEU", "CYS",
"MET", "LYS", "VAL", "LEU", "LEU", "LEU", "LEU", "ASN", "THR",
"ILE", "PRO", "LEU", "GLU", "GLY", "LEU", "PHE", "MET", "ARG",
"TYR", "ILE", "GLU", "LEU", "ALA", "ILE", "ARG", "ARG", "PHE",
"TYR", "GLN", "LEU", "THR", "LYS", "LEU", "LEU", "ASP", "ASN",
"LEU", "HIS", "ASP", "LEU", "VAL", "LYS", "GLN", "LEU", "HIS",
"LEU", "TYR", "CYS", "LEU", "ASN", "THR", "PHE", "ILE", "GLN",
"SER", "ARG", "ALA", "LEU", "SER", "VAL", "GLU", "PHE", "PRO",
"GLU", "MET", "MET", "SER", "GLU", "VAL", "ILE", "ALA", "ALA",
"GLN", "LEU", "PRO", "LYS", "ILE", "LEU", "ALA", "GLY", "MET",
"VAL", "LYS", "PRO", "LEU", "LEU", "PHE", "HIS", "LYS", "ASN",
"LEU", "ASP", "ASP", "ILE", "THR", "LEU", "ILE", "GLN", "TYR",
"SER", "TRP", "MET", "THR", "ILE", "PRO", "LEU", "GLU", "GLY",
"LEU", "ARG", "VAL", "LYS", "GLN", "LEU", "HIS", "LEU", "TYR",
"CYS", "LEU", "ASN", "THR", "PHE", "ILE", "GLN", "SER", "ARG",
"ALA", "LEU", "SER", "VAL", "GLU", "PHE", "PRO", "GLU", "MET",
"MET", "SER", "GLU", "VAL", "ILE", "ALA", "ALA", "GLN", "LEU",
"PRO", "LYS", "ILE", "LEU", "ALA", "GLY", "MET", "VAL", "LYS",
"PRO"
]
# Recall the keys have format RESNAME_RESNUMBER_CHAIN
resnames = [reskey.split("_")[0].strip() for reskey in residues]
resnames.sort()
prgr_residues.sort()
assert resnames == prgr_residues
# prgr.pdb has 2749 unique atoms.
atom_count = 0
for (_, atom_indices) in residues.iteritems():
atom_count += len(atom_indices)
print atom_count
assert atom_count == 2788
def test_get_lysine_charges(self):
"""
TestPDB: Test that lysine charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
lysine_charges = self.prgr_pdb.get_lysine_charges(res_list)
# prgr has 14 lysines.
print len(lysine_charges)
assert len(lysine_charges) == 14
for charge in lysine_charges:
# Lysine should be posistively charged
assert charge.positive
def test_get_arginine_charges(self):
"""
TestPDB: Test that arginine charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
arginine_charges = self.prgr_pdb.get_arginine_charges(res_list)
# prgr has 10 arginines
assert len(arginine_charges) == 10
for charge in arginine_charges:
# The guanidium in arginine should be positively charged.
assert charge.positive
def test_get_histidine_charges(self):
"""
TestPDB: Test that histidine charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
histidine_charges = self.prgr_pdb.get_histidine_charges(res_list)
# prgr has 7 arginines
assert len(histidine_charges) == 7
for charge in histidine_charges:
# The nitrogens pick up positive charges
assert charge.positive
def test_get_glutamic_acid_charges(self):
"""
TestPDB: Test that glutamic acid charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
glutamic_acid_charges = self.prgr_pdb.get_glutamic_acid_charges(
res_list)
assert len(glutamic_acid_charges) == 16
for charge in glutamic_acid_charges:
# The carboxyls get deprotonated.
assert not charge.positive
def test_get_aspartic_acid_charges(self):
"""
TestPDB: Test that aspartic acid charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
aspartic_acid_charges = self.prgr_pdb.get_aspartic_acid_charges(
res_list)
assert len(aspartic_acid_charges) == 9
for charge in aspartic_acid_charges:
# The carboxyls get deprotonated
assert not charge.positive
def test_assign_ligand_aromatics(self):
"""
TestPDB: Test that non-protein aromatic rings are assigned correctly.
"""
### 3ao4 comes from PDBBind-CN and contains some cruft in the PDB file:
### atoms without residues labelled. This triggered some problems with
### non-protein aromatics complaining.
# TODO(rbharath): Add a stub here.
_3ao4_protein = PDB()
_3ao4_protein_pdb = os.path.join(data_dir(), "3ao4_protein_hyd.pdb")
_3ao4_protein_pdbqt = os.path.join(data_dir(),
"3ao4_protein_hyd.pdbqt")
_3ao4_protein.load_from_files(_3ao4_protein_pdb, _3ao4_protein_pdbqt)
def test_remove_redundant_rings(self):
"""
TestPDB: Test that redundant rings are removed.
"""
# Recall that each ring is represented as a list of atom indices.
# Test that rings of length 0 are removed
assert remove_redundant_rings([[]]) == []
# Set that supersets are removed
assert (remove_redundant_rings([[1, 2, 3], [1, 3, 4, 5],
[1, 2, 3, 4, 5]]) == [[1, 2, 3],
[1, 3, 4, 5]])
# Ensure that duplicate rings are handled correctly (that is, only one
# copy of a duplicate ring should remain)
assert remove_redundant_rings([[1, 2, 3], [1, 3, 2]]) == [[1, 2, 3]]
def test_assign_protein_aromatics(self):
"""
TestPDB: Test that aromatic rings are assigned correctly.
"""
for name, protein in self.proteins:
# The proteins should have aromatic rings assigned already by
# load_from_files()
print "Processing aromatics for %s" % name
for aromatic in protein.aromatic_rings:
assert aromatic is not None
def test_get_phenylalanine_aromatics(self):
"""
TestPDB: Test that phenylalanine aromatic rings are retrieved.
"""
res_list = self.prgr_pdb.get_residues()
phenylalanine_aromatics = (
self.prgr_pdb.get_phenylalanine_aromatics(res_list))
# prgr has 13 phenylalanines, each of which has 1 aromatic ring.
assert len(phenylalanine_aromatics) == 13
for aromatic in phenylalanine_aromatics:
# The aromatic rings in phenylalanine have 6 elements each
assert len(aromatic.indices) == 6
def test_get_tyrosine_aromatics(self):
"""
TestPDB: Test that tyrosine aromatic rings are retrieved.
"""
# prgr has 10 tyrosines, each of which has 1 aromatic ring.
res_list = self.prgr_pdb.get_residues()
tyrosine_aromatics = self.prgr_pdb.get_tyrosine_aromatics(res_list)
assert len(tyrosine_aromatics) == 10
for aromatic in tyrosine_aromatics:
# The aromatic rings in tyrosine have 6 elements each
assert len(aromatic.indices) == 6
def test_get_histidine_aromatics(self):
"""
TestPDB: Test that histidine aromatic rings are retrieved.
"""
res_list = self.prgr_pdb.get_residues()
histidine_aromatics = self.prgr_pdb.get_histidine_aromatics(res_list)
# prgr has 7 histidines, each of which has 1 aromatic ring.
assert len(histidine_aromatics) == 7
for aromatic in histidine_aromatics:
# The aromatic rings in histidine have 6 elements each
print len(aromatic.indices)
assert len(aromatic.indices) == 5
def test_get_tryptophan_aromatics(self):
"""
TestPDB: Test that tryptophan aromatic rings are retrieved.
"""
res_list = self.prgr_pdb.get_residues()
tryptophan_aromatics = self.prgr_pdb.get_tryptophan_aromatics(res_list)
# prgr has 5 tryptophans, each of which has 2 aromatic ring.
print len(tryptophan_aromatics)
assert len(tryptophan_aromatics) == 10
num_five_rings, num_six_rings = 0, 0
for aromatic in tryptophan_aromatics:
# One aromatic ring in tryptophan hahas 6 elements each,
# while the other has 5 elements.
if len(aromatic.indices) == 6:
num_six_rings += 1
elif len(aromatic.indices) == 5:
num_five_rings += 1
assert num_six_rings == 5
assert num_five_rings == 5
def test_connected_atoms(self):
"""
TestPDB: Verifies that connected atom retrieval works.
"""
# Verify that no atoms are present when we start.
assert len(self.pdb.all_atoms.keys()) == 0
carbon_atom = Atom(element="C")
oxygen_atom = Atom(element="O")
hydrogen_atom = Atom(element="H")
self.pdb.add_new_atom(carbon_atom)
self.pdb.add_new_atom(oxygen_atom)
self.pdb.add_new_atom(hydrogen_atom)
# We want a carboxyl, so C connects O and H
carbon_atom.indices_of_atoms_connecting = [2, 3]
oxygen_atom.indices_of_atoms_connecting = [1]
hydrogen_atom.indices_of_atoms_connecting = [1]
connected_oxygens = self.pdb.connected_atoms(1, "O")
assert len(connected_oxygens) == 1
connected_hydrogens = self.pdb.connected_atoms(1, "H")
assert len(connected_hydrogens) == 1
def test_load_bonds_from_pdb(self):
"""
TestPDB: Verifies that bonds can be loaded from PDB.
"""
pdb = PDB()
# Test that we can load CO2
carbon_atom = Atom(element="C")
oxygen_atom_1 = Atom(element="O")
oxygen_atom_2 = Atom(element="O")
pdb.add_new_atom(carbon_atom)
pdb.add_new_atom(oxygen_atom_1)
pdb.add_new_atom(oxygen_atom_2)
lines = [
"CONECT 1 2 3 "
"CONECT 2 "
"CONECT 3 "
]
with tempfile.NamedTemporaryFile() as temp:
temp.write("\n".join(lines))
temp.flush()
pdb.load_bonds_from_pdb(temp.name)
assert len(carbon_atom.indices_of_atoms_connecting) == 2
assert len(oxygen_atom_1.indices_of_atoms_connecting) == 0
assert len(oxygen_atom_2.indices_of_atoms_connecting) == 0
def test_connected_heavy_atoms(self):
"""
TestPDB: Verifies retrieval of connected heavy atoms.
"""
# Verify that no atoms are present when we start.
assert len(self.pdb.all_atoms.keys()) == 0
carbon_atom = Atom(element="C")
oxygen_atom = Atom(element="O")
hydrogen_atom = Atom(element="H")
self.pdb.add_new_atom(carbon_atom)
self.pdb.add_new_atom(oxygen_atom)
self.pdb.add_new_atom(hydrogen_atom)
# We want a carboxyl, so C connects O and H
carbon_atom.indices_of_atoms_connecting = [2, 3]
oxygen_atom.indices_of_atoms_connecting = [1]
hydrogen_atom.indices_of_atoms_connecting = [1]
connected_heavy_atoms = self.pdb.connected_heavy_atoms(1)
assert len(connected_heavy_atoms) == 1
assert connected_heavy_atoms[0] == 2
def test_assign_non_protein_charges(self):
"""
TestPDB: Verify that charges are properly added to ligands.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# There should be 3 charged groups, two positive, one negative
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(
data_dir(), "ammonium_sulfate_hyd.pdbqt")
# Notice that load automatically identifies non-protein charges.
ammonium_sulfate_pdb.load_from_files(ammonium_sulfate_pdb_path,
ammonium_sulfate_pdbqt_path)
assert len(ammonium_sulfate_pdb.charges) == 3
num_pos, num_neg = 0, 0
for charge in ammonium_sulfate_pdb.charges:
if charge.positive:
num_pos += 1
else:
num_neg += 1
assert num_pos == 2
assert num_neg == 1
def test_metallic_charges(self):
"""
TestPDB: Verify that non-protein charges are assigned properly.
"""
# Test metallic ion charge.
magnesium_pdb = PDB()
magnesium_atom = Atom(element="MG",
coordinates=Point(coords=np.array([0, 0, 0])))
magnesium_pdb.add_new_non_protein_atom(magnesium_atom)
metallic_charges = magnesium_pdb.identify_metallic_charges()
assert len(metallic_charges) == 1
def test_nitrogen_charges(self):
"""
TestPDB: Verify that nitrogen groups are charged correctly.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# The labeling should pick up 2 charged nitrogen groups for two
# ammoniums.
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(
data_dir(), "ammonium_sulfate_hyd.pdbqt")
ammonium_sulfate_pdb.load_from_files(ammonium_sulfate_pdb_path,
ammonium_sulfate_pdbqt_path)
nitrogen_charges = ammonium_sulfate_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 2
assert nitrogen_charges[0].positive # Should be positive
assert nitrogen_charges[1].positive # Should be positive
# Test pyrrolidine (CH2)4NH. The nitrogen here should be sp3
# hybridized, so is likely to pick up an extra proton to its nitrogen
# at physiological pH.
pyrrolidine_pdb = PDB()
pyrrolidine_pdb_path = os.path.join(data_dir(), "pyrrolidine_hyd.pdb")
pyrrolidine_pdbqt_path = os.path.join(data_dir(),
"pyrrolidine_hyd.pdbqt")
pyrrolidine_pdb.load_from_files(pyrrolidine_pdb_path,
pyrrolidine_pdbqt_path)
nitrogen_charges = pyrrolidine_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 1
assert nitrogen_charges[0].positive # Should be positive
def test_carbon_charges(self):
"""
TestPDB: Verify that carbon groups are charged correctly.
"""
# Guanidine is positively charged at physiological pH
guanidine_pdb = PDB()
guanidine_pdb_path = os.path.join(data_dir(), "guanidine_hyd.pdb")
guanidine_pdbqt_path = os.path.join(data_dir(), "guanidine_hyd.pdbqt")
guanidine_pdb.load_from_files(guanidine_pdb_path, guanidine_pdbqt_path)
carbon_charges = guanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# sulfaguanidine contains a guanidine group that is likely to be
# positively protonated at physiological pH
sulfaguanidine_pdb = PDB()
sulfaguanidine_pdb_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdb")
sulfaguanidine_pdbqt_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdbqt")
sulfaguanidine_pdb.load_from_files(sulfaguanidine_pdb_path,
sulfaguanidine_pdbqt_path)
carbon_charges = sulfaguanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# Formic acid is a carboxylic acid, which should be negatively charged.
formic_acid_pdb = PDB()
formic_acid_pdb_path = os.path.join(data_dir(), "formic_acid_hyd.pdb")
formic_acid_pdbqt_path = os.path.join(data_dir(),
"formic_acid_hyd.pdbqt")
formic_acid_pdb.load_from_files(formic_acid_pdb_path,
formic_acid_pdbqt_path)
carbon_charges = formic_acid_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert not carbon_charges[0].positive # Should be negatively charged.
def test_phosphorus_charges(self):
"""
TestPDB: Verify that Phosphorus groups are charged correctly.
"""
# CID82671 contains a phosphate between two aromatic groups.
phosphate_pdb = PDB()
phosphate_pdb_path = os.path.join(data_dir(), "82671_hyd.pdb")
phosphate_pdbqt_path = os.path.join(data_dir(), "82671_hyd.pdb")
phosphate_pdb.load_from_files(phosphate_pdb_path, phosphate_pdbqt_path)
phosphorus_charges = phosphate_pdb.identify_phosphorus_charges()
assert len(phosphorus_charges) == 1
assert not phosphorus_charges[
0].positive # Should be negatively charged.
def test_sulfur_charges(self):
"""
TestPDB: Verify that sulfur groups are charged correctly.
"""
triflic_acid_pdb = PDB()
triflic_acid_pdb_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdb")
triflic_acid_pdbqt_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdbqt")
triflic_acid_pdb.load_from_files(triflic_acid_pdb_path,
triflic_acid_pdbqt_path)
sulfur_charges = (triflic_acid_pdb.identify_sulfur_charges())
assert len(sulfur_charges) == 1
assert not sulfur_charges[0].positive # Should be negatively charged.
def test_ligand_assign_aromatics(self):
"""
TestPDB: Verify that aromatic rings in ligands are identified.
"""
benzene_pdb = PDB()
benzene_pdb_path = os.path.join(data_dir(), "benzene_hyd.pdb")
benzene_pdbqt_path = os.path.join(data_dir(), "benzene_hyd.pdbqt")
benzene_pdb.load_from_files(benzene_pdb_path, benzene_pdbqt_path)
# A benzene should have exactly one aromatic ring.
print benzene_pdb.aromatic_rings
assert len(benzene_pdb.aromatic_rings) == 1
# The first 6 atoms in the benzene pdb form the aromatic ring.
assert (set(benzene_pdb.aromatic_rings[0].indices) == set(
[1, 2, 3, 4, 5, 6]))
def test_assign_secondary_structure(self):
"""
TestPDB: Verify that secondary structure is assigned meaningfully.
"""
# TODO(rbharath): This test is just a stub. Add a more realistic test
# that checks that nontrivial secondary structure is computed correctly
# here.
self.prgr_pdb.assign_secondary_structure()
def test_get_structure_dict(self):
"""
TestPDB: Verify that dict with rudimentary structure labels is generated.
TODO(rbharath): This is just a stub. Add some nontrivial tests here.
"""
structures = self.prgr_pdb.get_structure_dict()
print structures
print len(structures)
def featurize_fingerprint(pdb_directories, pickle_out):
"""Featurize all pdbs in provided directories."""
# Instantiate copy of binana vector
binana = Binana()
# See features/tests/nnscore_test.py:TestBinana.testComputeInputVector
# for derivation.
feature_len = binana.num_features()
feature_vectors = {}
for count, pdb_dir in enumerate(pdb_directories):
print "\nprocessing %d-th pdb %s" % (count, dir)
print "About to extract ligand and protein input files"
ligand_pdb, ligand_pdbqt = None, None
protein_pdb, protein_pdbqt = None, None
for f in os.listdir(pdb_dir):
if re.search("_ligand_hyd.pdb$", f):
ligand_pdb = f
elif re.search("_ligand_hyd.pdbqt$", f):
ligand_pdbqt = f
elif re.search("_protein_hyd.pdb$", f):
protein_pdb = f
elif re.search("_protein_hyd.pdbqt$", f):
protein_pdbqt = f
print "Extracted Input Files:"
print (ligand_pdb, ligand_pdbqt, protein_pdb, protein_pdbqt)
if (not ligand_pdb or not ligand_pdbqt or not protein_pdb or not
protein_pdbqt):
raise ValueError("Required files not present for %s" % pdb_dir)
ligand_pdb_path = os.path.join(pdb_dir, ligand_pdb)
ligand_pdbqt_path = os.path.join(pdb_dir, ligand_pdbqt)
protein_pdb_path = os.path.join(pdb_dir, protein_pdb)
protein_pdbqt_path = os.path.join(pdb_dir, protein_pdbqt)
print "About to load ligand from input files"
ligand_pdb_obj = PDB()
ligand_pdb_obj.load_from_files(ligand_pdb_path, ligand_pdbqt_path)
print "About to load protein from input files"
protein_pdb_obj = PDB()
protein_pdb_obj.load_from_files(protein_pdb_path, protein_pdbqt_path)
print "About to generate feature vector."
features = binana.compute_input_vector(ligand_pdb_obj,
protein_pdb_obj)
if len(features) != feature_len:
raise ValueError("Feature length incorrect on %s" % pdb_dir)
print "Feature vector generated correctly."
print "About to compute ligand smiles string."
ligand_mol = Chem.MolFromPDBFile(ligand_pdb_path)
# TODO(rbharath): Why does this fail sometimes?
if ligand_mol is None:
continue
smiles = Chem.MolToSmiles(ligand_mol)
print "About to compute sequence."
protein = md.load(protein_pdb_path)
seq = [r.name for r in protein.top.residues]
# Write the computed quantities
feature_vectors[pdb_dir] = (features, smiles, seq)
print "About to write pickle to " + pickle_out
with open(pickle_out, "wb") as f:
pickle.dump(feature_vectors, f)
class TestPDB(unittest.TestCase):
""""
Test PDB class.
"""
def setUp(self):
"""
Instantiate a dummy PDB file.
"""
self.temp_dir = tempfile.mkdtemp()
self.pdb = PDB()
_, self.pdb_filename = tempfile.mkstemp(suffix=".pdb",
dir=self.temp_dir)
self.prgr_pdb = PDB()
prgr_pdb_path = os.path.join(data_dir(), "prgr_hyd.pdb")
prgr_pdbqt_path = os.path.join(data_dir(), "prgr_hyd.pdbqt")
self.prgr_pdb.load_from_files(prgr_pdb_path, prgr_pdbqt_path)
self._1r5y_protein = PDB()
_1r5y_protein_pdb = os.path.join(data_dir(), "1r5y_protein_hyd.pdb")
_1r5y_protein_pdbqt = os.path.join(data_dir(), "1r5y_protein_hyd.pdbqt")
self._1r5y_protein.load_from_files(_1r5y_protein_pdb, _1r5y_protein_pdbqt)
self.proteins = [("prgr", self.prgr_pdb), ("1r5y", self._1r5y_protein)]
def tearDown(self):
"""
Delete temporary directory.
"""
shutil.rmtree(self.temp_dir)
def test_add_new_atom(self):
"""
TestPDB: Verifies that new atoms can be added.
"""
# Verify that no atoms are present when we start.
assert len(self.pdb.all_atoms.keys()) == 0
empty_atom = Atom()
self.pdb.add_new_atom(empty_atom)
# Verify that we now have one atom
assert len(self.pdb.all_atoms.keys()) == 1
def test_get_residues(self):
"""
TestPDB: Tests that all residues in PDB are identified.
"""
residues = self.prgr_pdb.get_residues()
# prgr.pdb has 280 unique residues
assert len(residues.keys()) == 280
prgr_residues = ["LEU", "ILE", "ASN", "LEU", "LEU", "MET", "SER",
"ILE", "GLU", "PRO", "ASP", "VAL", "ILE", "TYR", "ALA", "GLY", "HIS",
"ASP", "THR", "SER", "SER", "SER", "LEU", "LEU", "THR", "SER", "LEU",
"ASN", "GLN", "LEU", "GLY", "GLU", "ARG", "GLN", "LEU", "LEU", "SER",
"VAL", "VAL", "LYS", "TRP", "SER", "LYS", "SER", "LEU", "PRO", "GLY",
"PHE", "ARG", "LEU", "HIS", "ILE", "ASP", "ASP", "GLN", "ILE", "THR",
"LEU", "ILE", "GLN", "TYR", "SER", "TRP", "MET", "SER", "LEU", "MET",
"VAL", "PHE", "GLY", "LEU", "GLY", "TRP", "ARG", "SER", "TYR", "LYS",
"HIS", "VAL", "SER", "GLY", "GLN", "MET", "LEU", "TYR", "PHE", "ALA",
"PRO", "ASP", "LEU", "ILE", "LEU", "ASN", "GLU", "GLN", "ARG", "MET",
"LYS", "GLU", "PHE", "TYR", "SER", "LEU", "CYS", "LEU", "THR", "MET",
"TRP", "GLN", "ILE", "PRO", "GLN", "GLU", "PHE", "VAL", "LYS", "LEU",
"GLN", "VAL", "SER", "GLN", "GLU", "GLU", "PHE", "LEU", "CYS", "MET",
"LYS", "VAL", "LEU", "LEU", "LEU", "LEU", "ASN", "THR", "ILE", "PRO",
"LEU", "GLU", "GLY", "LEU", "PHE", "MET", "ARG", "TYR", "ILE", "GLU",
"LEU", "ALA", "ILE", "ARG", "ARG", "PHE", "TYR", "GLN", "LEU", "THR",
"LYS", "LEU", "LEU", "ASP", "ASN", "LEU", "HIS", "ASP", "LEU", "VAL",
"LYS", "GLN", "LEU", "HIS", "LEU", "TYR", "CYS", "LEU", "ASN", "THR",
"PHE", "ILE", "GLN", "SER", "ARG", "ALA", "LEU", "SER", "VAL", "GLU",
"PHE", "PRO", "GLU", "MET", "MET", "SER", "GLU", "VAL", "ILE", "ALA",
"ALA", "GLN", "LEU", "PRO", "LYS", "ILE", "LEU", "ALA", "GLY", "MET",
"VAL", "LYS", "PRO", "LEU", "LEU", "PHE", "HIS", "LYS", "ASN", "LEU",
"ASP", "ASP", "ILE", "THR", "LEU", "ILE", "GLN", "TYR", "SER", "TRP",
"MET", "THR", "ILE", "PRO", "LEU", "GLU", "GLY", "LEU", "ARG", "VAL",
"LYS", "GLN", "LEU", "HIS", "LEU", "TYR", "CYS", "LEU", "ASN", "THR",
"PHE", "ILE", "GLN", "SER", "ARG", "ALA", "LEU", "SER", "VAL", "GLU",
"PHE", "PRO", "GLU", "MET", "MET", "SER", "GLU", "VAL", "ILE", "ALA",
"ALA", "GLN", "LEU", "PRO", "LYS", "ILE", "LEU", "ALA", "GLY", "MET",
"VAL", "LYS", "PRO"]
# Recall the keys have format RESNAME_RESNUMBER_CHAIN
resnames = [reskey.split("_")[0].strip() for reskey in residues]
resnames.sort()
prgr_residues.sort()
assert resnames == prgr_residues
# prgr.pdb has 2749 unique atoms.
atom_count = 0
for (_, atom_indices) in residues.iteritems():
atom_count += len(atom_indices)
print atom_count
assert atom_count == 2788
def test_get_lysine_charges(self):
"""
TestPDB: Test that lysine charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
lysine_charges = self.prgr_pdb.get_lysine_charges(res_list)
# prgr has 14 lysines.
print len(lysine_charges)
assert len(lysine_charges) == 14
for charge in lysine_charges:
# Lysine should be posistively charged
assert charge.positive
def test_get_arginine_charges(self):
"""
TestPDB: Test that arginine charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
arginine_charges = self.prgr_pdb.get_arginine_charges(res_list)
# prgr has 10 arginines
assert len(arginine_charges) == 10
for charge in arginine_charges:
# The guanidium in arginine should be positively charged.
assert charge.positive
def test_get_histidine_charges(self):
"""
TestPDB: Test that histidine charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
histidine_charges = self.prgr_pdb.get_histidine_charges(res_list)
# prgr has 7 arginines
assert len(histidine_charges) == 7
for charge in histidine_charges:
# The nitrogens pick up positive charges
assert charge.positive
def test_get_glutamic_acid_charges(self):
"""
TestPDB: Test that glutamic acid charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
glutamic_acid_charges = self.prgr_pdb.get_glutamic_acid_charges(res_list)
assert len(glutamic_acid_charges) == 16
for charge in glutamic_acid_charges:
# The carboxyls get deprotonated.
assert not charge.positive
def test_get_aspartic_acid_charges(self):
"""
TestPDB: Test that aspartic acid charges are identified correctly.
"""
res_list = self.prgr_pdb.get_residues()
aspartic_acid_charges = self.prgr_pdb.get_aspartic_acid_charges(res_list)
assert len(aspartic_acid_charges) == 9
for charge in aspartic_acid_charges:
# The carboxyls get deprotonated
assert not charge.positive
def test_assign_ligand_aromatics(self):
"""
TestPDB: Test that non-protein aromatic rings are assigned correctly.
"""
### 3ao4 comes from PDBBind-CN and contains some cruft in the PDB file:
### atoms without residues labelled. This triggered some problems with
### non-protein aromatics complaining.
# TODO(rbharath): Add a stub here.
_3ao4_protein = PDB()
_3ao4_protein_pdb = os.path.join(data_dir(), "3ao4_protein_hyd.pdb")
_3ao4_protein_pdbqt = os.path.join(data_dir(), "3ao4_protein_hyd.pdbqt")
_3ao4_protein.load_from_files(_3ao4_protein_pdb, _3ao4_protein_pdbqt)
def test_remove_redundant_rings(self):
"""
TestPDB: Test that redundant rings are removed.
"""
# Recall that each ring is represented as a list of atom indices.
# Test that rings of length 0 are removed
assert remove_redundant_rings([[]]) == []
# Set that supersets are removed
assert (remove_redundant_rings([[1, 2, 3], [1, 3, 4, 5], [1, 2, 3, 4, 5]])
== [[1, 2, 3], [1, 3, 4, 5]])
# Ensure that duplicate rings are handled correctly (that is, only one
# copy of a duplicate ring should remain)
assert remove_redundant_rings([[1, 2, 3], [1, 3, 2]]) == [[1, 2, 3]]
def test_assign_protein_aromatics(self):
"""
TestPDB: Test that aromatic rings are assigned correctly.
"""
for name, protein in self.proteins:
# The proteins should have aromatic rings assigned already by
# load_from_files()
print "Processing aromatics for %s" % name
for aromatic in protein.aromatic_rings:
assert aromatic is not None
def test_get_phenylalanine_aromatics(self):
"""
TestPDB: Test that phenylalanine aromatic rings are retrieved.
"""
res_list = self.prgr_pdb.get_residues()
phenylalanine_aromatics = (
self.prgr_pdb.get_phenylalanine_aromatics(res_list))
# prgr has 13 phenylalanines, each of which has 1 aromatic ring.
assert len(phenylalanine_aromatics) == 13
for aromatic in phenylalanine_aromatics:
# The aromatic rings in phenylalanine have 6 elements each
assert len(aromatic.indices) == 6
def test_get_tyrosine_aromatics(self):
"""
TestPDB: Test that tyrosine aromatic rings are retrieved.
"""
# prgr has 10 tyrosines, each of which has 1 aromatic ring.
res_list = self.prgr_pdb.get_residues()
tyrosine_aromatics = self.prgr_pdb.get_tyrosine_aromatics(res_list)
assert len(tyrosine_aromatics) == 10
for aromatic in tyrosine_aromatics:
# The aromatic rings in tyrosine have 6 elements each
assert len(aromatic.indices) == 6
def test_get_histidine_aromatics(self):
"""
TestPDB: Test that histidine aromatic rings are retrieved.
"""
res_list = self.prgr_pdb.get_residues()
histidine_aromatics = self.prgr_pdb.get_histidine_aromatics(res_list)
# prgr has 7 histidines, each of which has 1 aromatic ring.
assert len(histidine_aromatics) == 7
for aromatic in histidine_aromatics:
# The aromatic rings in histidine have 6 elements each
print len(aromatic.indices)
assert len(aromatic.indices) == 5
def test_get_tryptophan_aromatics(self):
"""
TestPDB: Test that tryptophan aromatic rings are retrieved.
"""
res_list = self.prgr_pdb.get_residues()
tryptophan_aromatics = self.prgr_pdb.get_tryptophan_aromatics(res_list)
# prgr has 5 tryptophans, each of which has 2 aromatic ring.
print len(tryptophan_aromatics)
assert len(tryptophan_aromatics) == 10
num_five_rings, num_six_rings = 0, 0
for aromatic in tryptophan_aromatics:
# One aromatic ring in tryptophan hahas 6 elements each,
# while the other has 5 elements.
if len(aromatic.indices) == 6:
num_six_rings += 1
elif len(aromatic.indices) == 5:
num_five_rings += 1
assert num_six_rings == 5
assert num_five_rings == 5
def test_connected_atoms(self):
"""
TestPDB: Verifies that connected atom retrieval works.
"""
# Verify that no atoms are present when we start.
assert len(self.pdb.all_atoms.keys()) == 0
carbon_atom = Atom(element="C")
oxygen_atom = Atom(element="O")
hydrogen_atom = Atom(element="H")
self.pdb.add_new_atom(carbon_atom)
self.pdb.add_new_atom(oxygen_atom)
self.pdb.add_new_atom(hydrogen_atom)
# We want a carboxyl, so C connects O and H
carbon_atom.indices_of_atoms_connecting = [2, 3]
oxygen_atom.indices_of_atoms_connecting = [1]
hydrogen_atom.indices_of_atoms_connecting = [1]
connected_oxygens = self.pdb.connected_atoms(1, "O")
assert len(connected_oxygens) == 1
connected_hydrogens = self.pdb.connected_atoms(1, "H")
assert len(connected_hydrogens) == 1
def test_load_bonds_from_pdb(self):
"""
TestPDB: Verifies that bonds can be loaded from PDB.
"""
pdb = PDB()
# Test that we can load CO2
carbon_atom = Atom(element="C")
oxygen_atom_1 = Atom(element="O")
oxygen_atom_2 = Atom(element="O")
pdb.add_new_atom(carbon_atom)
pdb.add_new_atom(oxygen_atom_1)
pdb.add_new_atom(oxygen_atom_2)
lines = [
"CONECT 1 2 3 "
"CONECT 2 "
"CONECT 3 "
]
with tempfile.NamedTemporaryFile() as temp:
temp.write("\n".join(lines))
temp.flush()
pdb.load_bonds_from_pdb(temp.name)
assert len(carbon_atom.indices_of_atoms_connecting) == 2
assert len(oxygen_atom_1.indices_of_atoms_connecting) == 0
assert len(oxygen_atom_2.indices_of_atoms_connecting) == 0
def test_connected_heavy_atoms(self):
"""
TestPDB: Verifies retrieval of connected heavy atoms.
"""
# Verify that no atoms are present when we start.
assert len(self.pdb.all_atoms.keys()) == 0
carbon_atom = Atom(element="C")
oxygen_atom = Atom(element="O")
hydrogen_atom = Atom(element="H")
self.pdb.add_new_atom(carbon_atom)
self.pdb.add_new_atom(oxygen_atom)
self.pdb.add_new_atom(hydrogen_atom)
# We want a carboxyl, so C connects O and H
carbon_atom.indices_of_atoms_connecting = [2, 3]
oxygen_atom.indices_of_atoms_connecting = [1]
hydrogen_atom.indices_of_atoms_connecting = [1]
connected_heavy_atoms = self.pdb.connected_heavy_atoms(1)
assert len(connected_heavy_atoms) == 1
assert connected_heavy_atoms[0] == 2
def test_assign_non_protein_charges(self):
"""
TestPDB: Verify that charges are properly added to ligands.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# There should be 3 charged groups, two positive, one negative
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdbqt")
# Notice that load automatically identifies non-protein charges.
ammonium_sulfate_pdb.load_from_files(
ammonium_sulfate_pdb_path, ammonium_sulfate_pdbqt_path)
assert len(ammonium_sulfate_pdb.charges) == 3
num_pos, num_neg = 0, 0
for charge in ammonium_sulfate_pdb.charges:
if charge.positive:
num_pos += 1
else:
num_neg += 1
assert num_pos == 2
assert num_neg == 1
def test_metallic_charges(self):
"""
TestPDB: Verify that non-protein charges are assigned properly.
"""
# Test metallic ion charge.
magnesium_pdb = PDB()
magnesium_atom = Atom(element="MG",
coordinates=Point(coords=np.array([0,0,0])))
magnesium_pdb.add_new_non_protein_atom(magnesium_atom)
metallic_charges = magnesium_pdb.identify_metallic_charges()
assert len(metallic_charges) == 1
def test_nitrogen_charges(self):
"""
TestPDB: Verify that nitrogen groups are charged correctly.
"""
# Test ammonium sulfate: (NH4)+(NH4)+(SO4)(2-)
# The labeling should pick up 2 charged nitrogen groups for two
# ammoniums.
ammonium_sulfate_pdb = PDB()
ammonium_sulfate_pdb_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdb")
ammonium_sulfate_pdbqt_path = os.path.join(data_dir(),
"ammonium_sulfate_hyd.pdbqt")
ammonium_sulfate_pdb.load_from_files(
ammonium_sulfate_pdb_path, ammonium_sulfate_pdbqt_path)
nitrogen_charges = ammonium_sulfate_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 2
assert nitrogen_charges[0].positive # Should be positive
assert nitrogen_charges[1].positive # Should be positive
# Test pyrrolidine (CH2)4NH. The nitrogen here should be sp3
# hybridized, so is likely to pick up an extra proton to its nitrogen
# at physiological pH.
pyrrolidine_pdb = PDB()
pyrrolidine_pdb_path = os.path.join(data_dir(),
"pyrrolidine_hyd.pdb")
pyrrolidine_pdbqt_path = os.path.join(data_dir(),
"pyrrolidine_hyd.pdbqt")
pyrrolidine_pdb.load_from_files(pyrrolidine_pdb_path,
pyrrolidine_pdbqt_path)
nitrogen_charges = pyrrolidine_pdb.identify_nitrogen_charges()
assert len(nitrogen_charges) == 1
assert nitrogen_charges[0].positive # Should be positive
def test_carbon_charges(self):
"""
TestPDB: Verify that carbon groups are charged correctly.
"""
# Guanidine is positively charged at physiological pH
guanidine_pdb = PDB()
guanidine_pdb_path = os.path.join(data_dir(),
"guanidine_hyd.pdb")
guanidine_pdbqt_path = os.path.join(data_dir(),
"guanidine_hyd.pdbqt")
guanidine_pdb.load_from_files(
guanidine_pdb_path, guanidine_pdbqt_path)
carbon_charges = guanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# sulfaguanidine contains a guanidine group that is likely to be
# positively protonated at physiological pH
sulfaguanidine_pdb = PDB()
sulfaguanidine_pdb_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdb")
sulfaguanidine_pdbqt_path = os.path.join(data_dir(),
"sulfaguanidine_hyd.pdbqt")
sulfaguanidine_pdb.load_from_files(
sulfaguanidine_pdb_path, sulfaguanidine_pdbqt_path)
carbon_charges = sulfaguanidine_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert carbon_charges[0].positive # Should be positive
# Formic acid is a carboxylic acid, which should be negatively charged.
formic_acid_pdb = PDB()
formic_acid_pdb_path = os.path.join(data_dir(),
"formic_acid_hyd.pdb")
formic_acid_pdbqt_path = os.path.join(data_dir(),
"formic_acid_hyd.pdbqt")
formic_acid_pdb.load_from_files(
formic_acid_pdb_path, formic_acid_pdbqt_path)
carbon_charges = formic_acid_pdb.identify_carbon_charges()
assert len(carbon_charges) == 1
assert not carbon_charges[0].positive # Should be negatively charged.
def test_phosphorus_charges(self):
"""
TestPDB: Verify that Phosphorus groups are charged correctly.
"""
# CID82671 contains a phosphate between two aromatic groups.
phosphate_pdb = PDB()
phosphate_pdb_path = os.path.join(data_dir(),
"82671_hyd.pdb")
phosphate_pdbqt_path = os.path.join(data_dir(),
"82671_hyd.pdb")
phosphate_pdb.load_from_files(
phosphate_pdb_path, phosphate_pdbqt_path)
phosphorus_charges = phosphate_pdb.identify_phosphorus_charges()
assert len(phosphorus_charges) == 1
assert not phosphorus_charges[0].positive # Should be negatively charged.
def test_sulfur_charges(self):
"""
TestPDB: Verify that sulfur groups are charged correctly.
"""
triflic_acid_pdb = PDB()
triflic_acid_pdb_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdb")
triflic_acid_pdbqt_path = os.path.join(data_dir(),
"triflic_acid_hyd.pdbqt")
triflic_acid_pdb.load_from_files(
triflic_acid_pdb_path,
triflic_acid_pdbqt_path)
sulfur_charges = (
triflic_acid_pdb.identify_sulfur_charges())
assert len(sulfur_charges) == 1
assert not sulfur_charges[0].positive # Should be negatively charged.
def test_ligand_assign_aromatics(self):
"""
TestPDB: Verify that aromatic rings in ligands are identified.
"""
benzene_pdb = PDB()
benzene_pdb_path = os.path.join(data_dir(), "benzene_hyd.pdb")
benzene_pdbqt_path = os.path.join(data_dir(), "benzene_hyd.pdbqt")
benzene_pdb.load_from_files(benzene_pdb_path, benzene_pdbqt_path)
# A benzene should have exactly one aromatic ring.
print benzene_pdb.aromatic_rings
assert len(benzene_pdb.aromatic_rings) == 1
# The first 6 atoms in the benzene pdb form the aromatic ring.
assert (set(benzene_pdb.aromatic_rings[0].indices)
== set([1,2,3,4,5,6]))
def test_assign_secondary_structure(self):
"""
TestPDB: Verify that secondary structure is assigned meaningfully.
"""
# TODO(rbharath): This test is just a stub. Add a more realistic test
# that checks that nontrivial secondary structure is computed correctly
# here.
self.prgr_pdb.assign_secondary_structure()
def test_get_structure_dict(self):
"""
TestPDB: Verify that dict with rudimentary structure labels is generated.
TODO(rbharath): This is just a stub. Add some nontrivial tests here.
"""
structures = self.prgr_pdb.get_structure_dict()
print structures
print len(structures)