def test_read_hmm_evalue(self):
"""
Test that the hmm hits are well read, and returned only if evalue is < to the
given threshold.
"""
rep_name = "acba.007.p01.13"
replicon_id = 'ACBA.007.P01_13'
replicon_path = self.find_data(
os.path.join('Replicons', rep_name + '.fst'))
prot_file = self.find_data(
os.path.join('Proteins', replicon_id + '.prt'))
args = argparse.Namespace()
args.gembase = False
args.replicon = replicon_path
cfg = Config(args)
sequences_db = read_multi_prot_fasta(replicon_path)
replicon = next(sequences_db)
prot_db = ProdigalDB(replicon, cfg, prot_file=prot_file)
infile = self.find_data(
os.path.join("Results_Integron_Finder_{}".format(rep_name),
"tmp_{}".format(replicon_id),
"{}_intI.res".format(replicon_id)))
df1 = read_hmm(rep_name, prot_db, infile, cfg, evalue=1.95e-25)
exp1 = pd.DataFrame(data={
"Accession_number": rep_name,
"query_name": "intI_Cterm",
"ID_query": "-",
"ID_prot": "ACBA.007.P01_13_1",
"strand": 1,
"pos_beg": 55,
"pos_end": 1014,
"evalue": 1.9e-25
},
index=[0])
exp1 = exp1[[
"Accession_number", "query_name", "ID_query", "ID_prot", "strand",
"pos_beg", "pos_end", "evalue"
]]
pdt.assert_frame_equal(df1, exp1)
df2 = read_hmm(replicon_id, prot_db, infile, cfg, evalue=1.9e-25)
exp2 = pd.DataFrame(columns=[
"Accession_number", "query_name", "ID_query", "ID_prot", "strand",
"pos_beg", "pos_end", "evalue"
])
intcols = ["pos_beg", "pos_end", "strand"]
floatcol = ["evalue"]
exp2[intcols] = exp2[intcols].astype(int)
exp2[floatcol] = exp2[floatcol].astype(float)
pdt.assert_frame_equal(df2, exp2)
def test_read_multi(self):
"""
Test reading hmm results when there are multiple hits: 2 hits on the same protein: keep
only the one with the best evalue. 2 hits on 2 different proteins: keep the 2 proteins.
"""
replicon_id = 'ACBA.0917.00019'
contig_id = 'ACBA.0917.00019.0001'
result_dir_expected = self.find_data(
"Results_Integron_Finder_{}.gembase".format(replicon_id))
replicon_path = self.find_data(
os.path.join('Gembase', 'Replicons', replicon_id + '.fna'))
prot_file = os.path.join(result_dir_expected,
"tmp_{}".format(contig_id),
contig_id + '.prt')
args = argparse.Namespace()
args.gembase = True
args.replicon = replicon_path
cfg = Config(args)
sequences_db = read_multi_prot_fasta(replicon_path)
replicon = next(sequences_db)
prot_db = GembaseDB(replicon, cfg, prot_file=prot_file)
infile = self.find_data(
os.path.join('fictive_results',
"{}_intI_multi.res".format(contig_id)))
df = read_hmm(contig_id, prot_db, infile, cfg)
exp = pd.DataFrame(data={
"Accession_number": [contig_id] * 2,
"query_name": ["Phage_integrase"] * 2,
"ID_query": ["PF00589.16"] * 2,
"ID_prot":
["ACBA.0917.00019.i0001_00298", "ACBA.0917.00019.i0001_00338"],
"strand": [-1, -1],
"pos_beg": [311597, 350328],
"pos_end": [312631, 351248],
"evalue": [5.5e-66, 3.4e-51]
},
index=[0, 1])
exp = exp[[
"Accession_number", "query_name", "ID_query", "ID_prot", "strand",
"pos_beg", "pos_end", "evalue"
]]
pdt.assert_frame_equal(df, exp)
def test_read_hmm_gembase(self):
"""
Test that the hmm hits are well read, when the gembase format is used (.prt file is
provided, prodigal is not used to find the proteins).
"""
replicon_id = 'ACBA.0917.00019'
contig_id = 'ACBA.0917.00019.0001'
result_dir_expected = self.find_data(
"Results_Integron_Finder_{}.gembase".format(replicon_id))
replicon_path = self.find_data(
os.path.join('Gembase', 'Replicons', replicon_id + '.fna'))
prot_file = os.path.join(result_dir_expected,
"tmp_{}".format(contig_id),
contig_id + '.prt')
infile = os.path.join(result_dir_expected, "tmp_{}".format(contig_id),
"{}_intI.res".format(contig_id))
args = argparse.Namespace()
args.gembase = True
args.replicon = replicon_path
cfg = Config(args)
sequences_db = read_multi_prot_fasta(replicon_path)
replicon = next(sequences_db)
prot_db = GembaseDB(replicon, cfg, prot_file=prot_file)
df = read_hmm(contig_id, prot_db, infile, cfg)
exp = pd.DataFrame(data={
"Accession_number": contig_id,
"query_name": "intI_Cterm",
"ID_query": "-",
"ID_prot": "ACBA.0917.00019.i0001_00298",
"strand": -1,
"pos_beg": 311597,
"pos_end": 312631,
"evalue": 3.6e-25
},
index=[0])
exp = exp[[
"Accession_number", "query_name", "ID_query", "ID_prot", "strand",
"pos_beg", "pos_end", "evalue"
]]
pdt.assert_frame_equal(df, exp)
def test_read_hmm(self):
"""
Test that the hmm hits are well read
"""
rep_name = "acba.007.p01.13"
replicon_id = 'ACBA.007.P01_13'
replicon_path = self.find_data(
os.path.join('Replicons', rep_name + '.fst'))
prot_file = self.find_data(
os.path.join('Proteins', replicon_id + '.prt'))
args = argparse.Namespace()
args.gembase = False
args.replicon = replicon_path
cfg = Config(args)
sequences_db = read_multi_prot_fasta(replicon_path)
replicon = next(sequences_db)
prot_db = ProdigalDB(replicon, cfg, prot_file=prot_file)
infile = self.find_data(
os.path.join("Results_Integron_Finder_{}".format(rep_name),
"tmp_{}".format(replicon_id),
"{}_intI.res".format(replicon_id)))
df = read_hmm(rep_name, prot_db, infile, cfg)
exp = pd.DataFrame(data={
"Accession_number": rep_name,
"query_name": "intI_Cterm",
"ID_query": "-",
"ID_prot": "ACBA.007.P01_13_1",
"strand": 1,
"pos_beg": 55,
"pos_end": 1014,
"evalue": 1.9e-25
},
index=[0])
exp = exp[[
"Accession_number", "query_name", "ID_query", "ID_prot", "strand",
"pos_beg", "pos_end", "evalue"
]]
pdt.assert_frame_equal(df, exp)
def test_read_hmm_cov2(self):
"""
Test that the hmm hits are well read, it returns only the hits with coverage >
given threshold
"""
rep_name = "acba.007.p01.13"
replicon_id = 'ACBA.007.P01_13'
replicon_path = self.find_data(
os.path.join('Replicons', rep_name + '.fst'))
prot_file = self.find_data(
os.path.join('Proteins', replicon_id + '.prt'))
args = argparse.Namespace()
args.gembase = False
args.replicon = replicon_path
cfg = Config(args)
sequences_db = read_multi_prot_fasta(replicon_path)
replicon = next(sequences_db)
prot_db = ProdigalDB(replicon, cfg, prot_file=prot_file)
infile = self.find_data(
os.path.join("fictive_results", "{}_intI.res".format(replicon_id)))
df1 = read_hmm(rep_name, prot_db, infile, cfg, coverage=0.7)
exp1 = pd.DataFrame(data={
"Accession_number": [rep_name] * 2,
"query_name": ["intI_Cterm"] * 2,
"ID_query": ["-", "-"],
"ID_prot": ["ACBA.007.P01_13_1", "ACBA.007.P01_13_2"],
"strand": [1, -1],
"pos_beg": [55, 905],
"pos_end": [1014, 1609],
"evalue": [1.9e-25, 1e-3]
},
index=[0, 1])
exp1 = exp1[[
"Accession_number", "query_name", "ID_query", "ID_prot", "strand",
"pos_beg", "pos_end", "evalue"
]]
pdt.assert_frame_equal(df1, exp1)
def test_read_empty(self):
"""
Test that when there are no hits in the hmm result file, it returns an empty
dataframe, without error.
"""
rep_name = "acba.007.p01.13"
replicon_id = 'ACBA.007.P01_13'
replicon_path = self.find_data(
os.path.join('Replicons', rep_name + '.fst'))
prot_file = self.find_data(
os.path.join('Proteins', replicon_id + '.prt'))
args = argparse.Namespace()
args.gembase = False
args.replicon = replicon_path
cfg = Config(args)
sequences_db = read_multi_prot_fasta(replicon_path)
replicon = next(sequences_db)
prot_db = ProdigalDB(replicon, cfg, prot_file=prot_file)
infile = self.find_data(
os.path.join("fictive_results",
"{}_intI-empty.res".format(replicon_id)))
df = read_hmm(rep_name, prot_db, infile, cfg)
exp = pd.DataFrame(columns=[
"Accession_number", "query_name", "ID_query", "ID_prot", "strand",
"pos_beg", "pos_end", "evalue"
])
intcols = ["pos_beg", "pos_end", "strand"]
floatcol = ["evalue"]
exp[intcols] = exp[intcols].astype(int)
exp[floatcol] = exp[floatcol].astype(float)
pdt.assert_frame_equal(df, exp)
