def test_extra_fields():
path = join(data_dir("structure"), "1l2y.cif")
pdbx_file = pdbx.PDBxFile.read(path)
stack1 = pdbx.get_structure(
pdbx_file, extra_fields=["atom_id", "b_factor", "occupancy", "charge"])
pdbx_file = pdbx.PDBxFile()
pdbx.set_structure(pdbx_file, stack1, data_block="test")
stack2 = pdbx.get_structure(
pdbx_file, extra_fields=["atom_id", "b_factor", "occupancy", "charge"])
assert stack1 == stack2
path = join(data_dir("structure"), "1l2y.cif")
pdbx_file = pdbx.PDBxFile.read(path)
stack1 = pdbx.get_structure(
pdbx_file, extra_fields=["atom_id", "b_factor", "occupancy", "charge"])
pdbx_file = pdbx.PDBxFile()
pdbx.set_structure(pdbx_file, stack1, data_block="test")
stack2 = pdbx.get_structure(
pdbx_file, extra_fields=["atom_id", "b_factor", "occupancy", "charge"])
assert stack1.ins_code.tolist() == stack2.ins_code.tolist()
assert stack1.atom_id.tolist() == stack2.atom_id.tolist()
assert stack1.b_factor.tolist() == approx(stack2.b_factor.tolist())
assert stack1.occupancy.tolist() == approx(stack2.occupancy.tolist())
assert stack1.charge.tolist() == stack2.charge.tolist()
assert stack1 == stack2
def test_conversion(path, model):
pdbx_file = pdbx.PDBxFile.read(path)
try:
array1 = pdbx.get_structure(pdbx_file, model=model)
except biotite.InvalidFileError:
if model is None:
# The file cannot be parsed into an AtomArrayStack,
# as the models contain different numbers of atoms
# -> skip this test case
return
else:
raise
pdbx_file = pdbx.PDBxFile()
pdbx.set_structure(pdbx_file, array1, data_block="test")
# Remove one optional auth section in label to test fallback to label fields
atom_cat = pdbx_file.get_category("atom_site", "test")
atom_cat.pop("auth_atom_id")
pdbx_file.set_category("atom_site", atom_cat, "test")
array2 = pdbx.get_structure(pdbx_file, model=model)
assert array1.array_length() > 0
if array1.box is not None:
assert np.allclose(array1.box, array2.box)
assert array1.bonds == array2.bonds
for category in array1.get_annotation_categories():
assert (array1.get_annotation(category).tolist() ==
array2.get_annotation(category).tolist())
assert array1.coord.tolist() == array2.coord.tolist()
def test_conversion(path, model):
pdbx_file = pdbx.PDBxFile.read(path)
try:
array1 = pdbx.get_structure(pdbx_file, model=model)
except biotite.InvalidFileError:
if model is None:
# The file cannot be parsed into an AtomArrayStack,
# as the models contain different numbers of atoms
# -> skip this test case
return
else:
raise
pdbx_file = pdbx.PDBxFile()
pdbx.set_structure(pdbx_file, array1, data_block="test")
array2 = pdbx.get_structure(pdbx_file, model=model)
if array1.box is not None:
assert np.allclose(array1.box, array2.box)
assert array1.bonds == array2.bonds
for category in array1.get_annotation_categories():
assert array1.get_annotation(category).tolist() == \
array2.get_annotation(category).tolist()
assert array1.coord.tolist() == array2.coord.tolist()
def test_extra_fields():
path = join(data_dir, "1l2y.cif")
pdbx_file = pdbx.PDBxFile()
pdbx_file.read(path)
stack1 = pdbx.get_structure(pdbx_file, extra_fields=["atom_id","b_factor",
"occupancy","charge"])
pdbx_file = pdbx.PDBxFile()
pdbx.set_structure(pdbx_file, stack1, data_block="test")
stack2 = pdbx.get_structure(pdbx_file, extra_fields=["atom_id","b_factor",
"occupancy","charge"])
assert stack1 == stack2
def test_conversion(path, single_model):
model = 1 if single_model else None
pdbx_file = pdbx.PDBxFile()
pdbx_file.read(path)
array1 = pdbx.get_structure(pdbx_file, model=model)
pdbx_file = pdbx.PDBxFile()
pdbx.set_structure(pdbx_file, array1, data_block="test")
array2 = pdbx.get_structure(pdbx_file, model=model)
for category in array1.get_annotation_categories():
assert array1.get_annotation(category).tolist() == \
array2.get_annotation(category).tolist()
assert array1.coord.tolist() == array2.coord.tolist()
ku_file = biotite.temp_file("ku.cif")
# Download and parse structure files
file = rcsb.fetch("1JEY", "mmtf", biotite.temp_dir())
ku_dna = strucio.load_structure(file)
file = rcsb.fetch("1JEQ", "mmtf", biotite.temp_dir())
ku = strucio.load_structure(file)
# Remove DNA and water
ku_dna = ku_dna[(ku_dna.chain_id == "A") | (ku_dna.chain_id == "B")]
ku_dna = ku_dna[~struc.filter_solvent(ku_dna)]
ku = ku[~struc.filter_solvent(ku)]
# The structures have a differing amount of atoms missing
# at the the start and end of the structure
# -> Find common structure
ku_dna_common = ku_dna[struc.filter_intersection(ku_dna, ku)]
ku_common = ku[struc.filter_intersection(ku, ku_dna)]
# Superimpose
ku_superimposed, transformation = struc.superimpose(
ku_dna_common, ku_common, (ku_common.atom_name == "CA"))
# We do not want the cropped structures
# -> apply superimposition on structures before intersection filtering
ku_superimposed = struc.superimpose_apply(ku, transformation)
# Write PDBx files as input for PyMOL
cif_file = pdbx.PDBxFile()
pdbx.set_structure(cif_file, ku_dna, data_block="ku_dna")
cif_file.write(ku_dna_file)
cif_file = pdbx.PDBxFile()
pdbx.set_structure(cif_file, ku_superimposed, data_block="ku")
cif_file.write(ku_file)
# Visualization with PyMOL...
# biotite_static_image = ku_superimposition.png
chain_combinations = list(combinations(chain_list, 2))
query = 'UPDATE cif SET chain_count = ?, combination_count = ? WHERE pdb_id = ?', (
len(chain_list), len(chain_combinations), pdb_id)
sql_queries.append(query)
for comb in chain_combinations:
dimer = model[np.logical_or(model.chain_id == comb[0],
model.chain_id == comb[1])]
sasa_result = pifaceFunctions.sasa(dimer, comb[0], comb[1])
if sasa_result[0] + sasa_result[1] - sasa_result[2] > args.sasa:
pdbxFile = PDBxFile()
set_structure(pdbxFile,
dimer,
data_block=pdb_id + '_' + comb[0] + '_' +
comb[1])
pdbxFile.write("%s_%s_%s.cif" % (pdb_id, comb[0], comb[1]))
dimer_id = pdb_id + '_' + comb[0] + '_' + comb[1]
query = 'INSERT INTO dimers (dimer_id, pdb_id, chain_1, chain_2, chain_1_sasa, chain_2_sasa, dimer_sasa, sasa_diff, insert_time) VALUES(?,?,?,?,?,?,?,?,?)', (
dimer_id, pdb_id, comb[0], comb[1],
float(sasa_result[0]), float(sasa_result[1]),
float(sasa_result[2]),
float(sasa_result[0] + sasa_result[1] -
sasa_result[2]),
datetime.strftime(datetime.now(), "%Y-%m-%d %H:%M:%S"))
sql_queries.append(query)
with open("%s_%s_%s.cif" % (pdb_id, comb[0], comb[1]),
file["audit_author"] = {
"name": ["Doe, Jane", "Doe, John"],
"pdbx_ordinal": ["1", "2"]
}
########################################################################
# In most applications only the structure itself
# (stored in the *atom_site* category) is relevant.
# :func:`get_structure()` and :func:`set_structure()` are convenience
# functions that are used to convert the
# *atom_site* category into an atom array (stack) and vice versa.
tc5b = pdbx.get_structure(file)
# Do some fancy stuff
pdbx.set_structure(file, tc5b)
########################################################################
# :func:`get_structure()` creates automatically an
# :class:`AtomArrayStack`, even if the file actually contains only a
# single model.
# If you would like to have an :class:`AtomArray` instead, you have to
# specifiy the :obj:`model` parameter.
#
# .. currentmodule:: biotite.structure.io.mmtf
#
# If you want to parse a large batch of structure files or you have to
# load very large structure files, the usage of PDB or mmCIF files might
# be too slow for your requirements. In this case you probably might
# want to use MMTF files.
# MMTF files describe structures just like PDB and mmCIF files,
cif_file["audit_author"] = {
"name": ["Doe, Jane", "Doe, John"],
"pdbx_ordinal": ["1", "2"]
}
########################################################################
# In most applications only the structure itself
# (stored in the *atom_site* category) is relevant.
# :func:`get_structure()` and :func:`set_structure()` are convenience
# functions that are used to convert the
# ``atom_site`` category into an atom array (stack) and vice versa.
tc5b = pdbx.get_structure(cif_file)
# Do some fancy stuff
pdbx.set_structure(cif_file, tc5b)
########################################################################
# :func:`get_structure()` creates automatically an
# :class:`AtomArrayStack`, even if the file actually contains only a
# single model.
# If you would like to have an :class:`AtomArray` instead, you have to
# specifiy the :obj:`model` parameter.
#
# .. currentmodule:: biotite.structure.io.mmtf
#
# If you want to parse a large batch of structure files or you have to
# load very large structure files, the usage of PDB or mmCIF files might
# be too slow for your requirements.
# In this case you probably might want to use MMTF files.
# MMTF files describe structures just like PDB and mmCIF files,