def test_annot_in0(self):
"""
Test func_annot when the integron is a in0: only an integrase. There are no proteins
to annotate, but the _subseqprot.tmp file already exists (not deleted in a last run
for example...)
"""
# create empty _subseqprot.tmp file (must be deleted by func_annot)
open(os.path.join(self.tmp_dir, "{}_subseqprot.tmp".format(self.replicon.id)), "w").close()
# Create integron
integron1 = Integron(self.replicon, self.cfg)
integrons = [integron1]
# Add integrase
integron1.add_integrase(55, 1014, "ACBA.007.P01_13_1", 1, 1.9e-25, "intersection_tyr_intI")
# check proteins before annotation
proteins = pd.DataFrame(columns=["pos_beg", "pos_end", "strand",
"evalue", "type_elt", "model",
"distance_2attC", "annotation"])
proteins = proteins.astype(dtype={"pos_beg": "int", "pos_end": "int", "strand": "int",
"evalue": "float", "type_elt": "str", "model": "str",
"distance_2attC": "float", "annotation": "str"})
pdt.assert_frame_equal(proteins, integron1.proteins)
# Annotate proteins
func_annot(integrons, self.replicon, self.prot_db, self.hmm_files, self.cfg, self.tmp_dir)
# Check that all files generated are as expected
files_created = [f for f in glob.glob(os.path.join(self.tmp_dir, "*")) if os.path.isfile(f)]
exp_files = ["{}{}".format(self.replicon.id, suffix) for suffix in ("_intI_table.res",
"_phage_int_table.res",
"_intI.res",
"_phage_int.res")]
exp_files = [os.path.join(self.tmp_dir, file) for file in exp_files]
self.assertEqual(set(exp_files), set(files_created))
# check proteins after annotation
pdt.assert_frame_equal(proteins, integron1.proteins)
def test_has_integrase(self):
replicon = SeqRecord(Seq.Seq(''), id='foo', name='bar')
integron = Integron(replicon, self.cfg)
self.assertFalse(integron.has_integrase())
replicon = SeqRecord(Seq.Seq(''), id='just_one_integrase', name='bar')
just_one_integrase = Integron(replicon, self.cfg)
just_one_integrase.add_integrase(10, 100, 'foo', 1, 1e-2,
"intersection_tyr_intI")
self.assertTrue(just_one_integrase.has_integrase())
def test_add_integrase(self):
replicon_name = "acba.007.p01.13"
replicon_path = self.find_data(os.path.join('Replicons', replicon_name + '.fst'))
topologies = Topology('lin')
with FastaIterator(replicon_path) as sequences_db:
sequences_db.topologies = topologies
replicon = next(sequences_db)
data_integrase = {"pos_beg": 55,
"pos_end": 1014,
"strand": 1,
"evalue": 1.900000e-25,
"type_elt": "protein",
"annotation": "intI",
"model": "intersection_tyr_intI",
"distance_2attC": np.nan}
id_int = "ACBA.007.P01_13_1"
df = pd.DataFrame(data_integrase,
columns=self.columns,
index=[id_int])
df = df.astype(dtype=self.dtype)
integron = Integron(replicon, self.cfg)
integron.add_integrase(data_integrase["pos_beg"],
data_integrase["pos_end"],
id_int,
data_integrase["strand"],
data_integrase["evalue"],
data_integrase["model"]
)
pdt.assert_frame_equal(df, integron.integrase)
with self.assertRaises(RuntimeError) as ctx:
integron.add_integrase(data_integrase["pos_beg"],
data_integrase["pos_end"],
id_int,
data_integrase["strand"],
data_integrase["evalue"],
data_integrase["model"]
)
self.assertEqual(str(ctx.exception), "add_integrase should be called once.")
def test_type(self):
replicon = SeqRecord(Seq.Seq(''), id='foo')
no_integrase = Integron(replicon, self.cfg)
self.assertIsNone(no_integrase.type())
replicon = SeqRecord(Seq.Seq(''), id='just_one_integrase')
just_one_integrase = Integron(replicon, self.cfg)
just_one_integrase.add_integrase(10,
100,
'foo',
1,
1e-2,
"intersection_tyr_intI")
self.assertEqual(just_one_integrase.type(), "In0")
replicon = SeqRecord(Seq.Seq(''), id='just_one_attC')
just_one_attC = Integron(replicon, self.cfg)
just_one_attC.add_attC(10,
100,
1,
1e-2,
"intersection_tyr_intI")
self.assertEqual(just_one_attC.type(), "CALIN")
replicon = SeqRecord(Seq.Seq(''), id='one_integrase_one_attC')
one_integrase_one_attC = Integron(replicon, self.cfg)
one_integrase_one_attC.add_integrase(10,
100,
'foo',
1,
1e-2,
"intersection_tyr_intI")
one_integrase_one_attC.add_attC(10,
100,
1,
1e-2,
"intersection_tyr_intI")
self.assertEqual(one_integrase_one_attC.type(), "complete")
def test_annot_multi(self):
"""
Test func_annot when there are 4 integrons:
- 1 calin with 4 proteins, 2 having a resfam annotation
- 1 calin with 2 proteins, none having a resfam annotation
- 1 in0
- 1 complete with 4 proteins, 3 having a resfam annotation
"""
# resfam pour: 16, 13, 3, 12
# Create integron in0
integron1 = Integron(self.replicon.name, self.cfg)
integron1.add_integrase(56, 1014, "ACBA.007.P01_13_1", 1, 1.9e-25, "intersection_tyr_intI")
# Create integron CALIN with resfam proteins
integron2 = Integron(self.replicon, self.cfg)
integron2.add_attC(7400, 7650, -1, 7e-9, "attc_4")
integron2.add_attC(8600, 8650, -1, 7e-4, "attc_4")
integron2.add_attC(10200, 10400, -1, 7e-7, "attc_4")
integron2.add_attC(10800, 10900, -1, 7e-7, "attc_4")
integron2.add_proteins(self.prot_db)
# Create integron CALIN without any resfam proteins
integron3 = Integron(self.replicon, self.cfg)
integron3.add_attC(4320, 4400, -1, 7e-9, "attc_4")
integron3.add_proteins(self.prot_db)
# Create complete integron
integron4 = Integron(self.replicon, self.cfg)
integron4.add_attC(17825, 17884, -1, 7e-9, "attc_4")
integron4.add_attC(19080, 19149, -1, 7e-4, "attc_4")
integron4.add_attC(19618, 19726, -1, 7e-7, "attc_4")
integron4.add_integrase(16542, 17381, "ACBA.007.P01_13_19", -1, 1.9e-25, "intersection_tyr_intI")
integron4.add_proteins(self.prot_db)
integrons = [integron1, integron2, integron3, integron4]
# Create dataframes for expected proteins before annotation
proteins1 = pd.DataFrame(columns=["pos_beg", "pos_end", "strand",
"evalue", "type_elt", "model",
"distance_2attC", "annotation"])
proteins1 = proteins1.astype(dtype={"pos_beg": "int", "pos_end": "int", "strand": "int",
"evalue": "float", "type_elt": "str", "model": "str",
"distance_2attC": "float", "annotation": "str"})
proteins1 = proteins1[["pos_beg", "pos_end", "strand", "evalue", "type_elt",
"model", "distance_2attC", "annotation"]]
proteins1 = proteins1.astype(dtype=self.prot_dtype)
proteins2 = pd.DataFrame({"pos_beg": [7088, 7710, 8650, 10524],
"pos_end": [7351, 8594, 10125, 11699],
"strand": [1, -1, -1, -1],
"evalue": [np.nan] * 4,
"type_elt": ["protein"] * 4,
"model": ["NA"] * 4,
"distance_2attC": [np.nan] * 4,
"annotation": ["protein"] * 4},
index=["ACBA.007.P01_13_11", "ACBA.007.P01_13_12",
"ACBA.007.P01_13_13", "ACBA.007.P01_13_14"])
proteins2 = proteins2[["pos_beg", "pos_end", "strand", "evalue", "type_elt",
"model", "distance_2attC", "annotation"]]
proteins2 = proteins2.astype(dtype=self.prot_dtype)
proteins3 = pd.DataFrame({"pos_beg": [3546, 4380],
"pos_end": [4313, 4721],
"strand": [1, 1],
"evalue": [np.nan] * 2,
"type_elt": ["protein"] * 2,
"model": ["NA"] * 2,
"distance_2attC": [np.nan] * 2,
"annotation": ["protein"] * 2},
index=["ACBA.007.P01_13_6", "ACBA.007.P01_13_7"])
proteins3 = proteins3[["pos_beg", "pos_end", "strand", "evalue", "type_elt",
"model", "distance_2attC", "annotation"]]
proteins3 = proteins3.astype(dtype=self.prot_dtype)
proteins4 = pd.DataFrame({"pos_beg": [17375, 17886, 19090, 19721],
"pos_end": [17722, 18665, 19749, 20254],
"strand": [-1] * 4,
"evalue": [np.nan] * 4,
"type_elt": ["protein"] * 4,
"model": ["NA"] * 4,
"distance_2attC": [np.nan] * 4,
"annotation": ["protein"] * 4},
index=["ACBA.007.P01_13_20", "ACBA.007.P01_13_21",
"ACBA.007.P01_13_22", "ACBA.007.P01_13_23"])
proteins4 = proteins4[["pos_beg", "pos_end", "strand", "evalue", "type_elt",
"model", "distance_2attC", "annotation"]]
proteins4 = proteins4.astype(dtype=self.prot_dtype)
# Check proteins before annotation
expected_proteins = [proteins1, proteins2, proteins3, proteins4]
for inte, exp_prot in zip(integrons, expected_proteins):
# we need to sort the dataframe
# as protein file is parse using biopython and index
# the order os sequences is not guarantee
pdt.assert_frame_equal(inte.proteins.sort_index(), exp_prot.sort_index())
# Annotate proteins with evalue threshold
func_annot(integrons, self.replicon, self.prot_db, self.hmm_files, self.cfg, self.tmp_dir, evalue=1e-32)
# Check that all files generated are as expected
files_created = [f for f in glob.glob(os.path.join(self.tmp_dir, "*")) if os.path.isfile(f)]
self.assertEqual(set(self.exp_files), set(files_created))
# Check that annotated proteins are as expected
proteins2.loc["ACBA.007.P01_13_13"] = [8650, 10125, -1, 2.4e-86, "protein",
"RF0007", np.nan, "ABC_efflux"]
proteins4.loc["ACBA.007.P01_13_21"] = [17886, 18665, -1, 7.4e-168, "protein",
"RF0027", np.nan, "ANT3"]
proteins4.loc["ACBA.007.P01_13_23"] = [19721, 20254, -1, 6.2e-110, "protein",
"RF0003", np.nan, "AAC3-I"]
for inte, prots in zip(integrons, expected_proteins):
# we need to sort the dataframe
# as protein file is parse using biopython and index
# the order os sequences is not guarantee
pdt.assert_frame_equal(inte.proteins.sort_index(), prots.sort_index())
# Annotate proteins with default evalue (1 more annotation)
with self.catch_io(out=True):
func_annot(integrons, self.replicon, self.prot_db, self.hmm_files, self.cfg, self.tmp_dir)
proteins4.loc["ACBA.007.P01_13_20"] = [17375, 17722, -1, 4.5e-31, "protein",
"RF0066", np.nan, "emrE"]
for inte, prots in zip(integrons, expected_proteins):
pdt.assert_frame_equal(inte.proteins.sort_index(), prots.sort_index())
# Annotate proteins with lower coverage threshold (1 more annotation)
with self.catch_io(out=True):
func_annot(integrons, self.replicon, self.prot_db, self.hmm_files, self.cfg, self.tmp_dir, coverage=0.4)
proteins2.loc["ACBA.007.P01_13_12"] = [7710, 8594, -1, 1.6e-5, "protein",
"RF0033", np.nan, "APH3"]
for inte, prots in zip(integrons, expected_proteins):
pdt.assert_frame_equal(inte.proteins.sort_index(), prots.sort_index())