def test_appending_dfs(self):
s = SNPs()
s._snps = self.create_snp_df(rsid=["rs1"],
chrom=["1"],
pos=[1],
genotype=["AA"])
s._duplicate = self.create_snp_df(rsid=["rs1"],
chrom=["1"],
pos=[1],
genotype=["AA"])
s._discrepant_XY = self.create_snp_df(rsid=["rs1"],
chrom=["1"],
pos=[1],
genotype=["AA"])
s.merge([s])
df = self.create_snp_df(rsid=["rs1", "rs1"],
chrom=["1", "1"],
pos=[1, 1],
genotype=["AA", "AA"])
pd.testing.assert_frame_equal(s.duplicate, df, check_exact=True)
pd.testing.assert_frame_equal(s.discrepant_XY, df, check_exact=True)
pd.testing.assert_frame_equal(s.heterozygous_MT,
get_empty_snps_dataframe(),
check_exact=True)
pd.testing.assert_frame_equal(s.discrepant_vcf_position,
get_empty_snps_dataframe(),
check_exact=True)
def test_merge_invalid_file(self):
s = SNPs()
results = s.merge(
[SNPs("tests/input/GRCh37.csv"), SNPs("tests/input/empty.txt")]
)
pd.testing.assert_frame_equal(s.snps, self.snps_GRCh37(), check_exact=True)
self.assert_results(results, [{"merged": True}, {}])
def test_merge_list(self):
s = SNPs()
results = s.merge(
[SNPs("tests/input/GRCh37.csv"),
SNPs("tests/input/GRCh37.csv")])
pd.testing.assert_frame_equal(s.snps,
self.snps_GRCh37(),
check_exact=True)
self.assertEqual(s.source, "generic, generic")
self.assertListEqual(s._source, ["generic", "generic"])
self.assert_results(
results,
[
{
"merged": True
},
{
"merged":
True,
"common_rsids":
pd.Index(
[
"rs3094315", "rs2500347", "rsIndelTest",
"rs11928389"
],
name="rsid",
),
},
],
)
def test_merge_non_existent_file(self):
s = SNPs()
results = s.merge(
[SNPs("tests/input/non_existent_file.csv"), SNPs("tests/input/GRCh37.csv")]
)
pd.testing.assert_frame_equal(s.snps, self.snps_GRCh37(), check_exact=True)
self.assert_results(results, [{}, {"merged": True}])
def f():
s = SNPs("tests/input/NCBI36.csv")
results = s.merge([SNPs("tests/input/GRCh37.csv")], remap=False)
df = s.discrepant_snps
self.assertEqual(len(df), 4)
pd.testing.assert_index_equal(
df.index,
results[0]["discrepant_position_rsids"],
check_exact=True,
check_names=True,
)
def test_merge_exceed_discrepant_genotypes_threshold(self):
s1 = SNPs("tests/input/generic.csv")
s2 = SNPs("tests/input/generic.csv")
s2._snps.loc["rs1", "genotype"] = "CC"
results = s1.merge([s2], discrepant_genotypes_threshold=0)
self.assertEqual(len(s1.discrepant_merge_positions), 0)
self.assertEqual(len(s1.discrepant_merge_genotypes), 0)
self.assertEqual(len(s1.discrepant_merge_positions_genotypes), 0)
pd.testing.assert_frame_equal(s1.snps, self.generic_snps(), check_exact=True)
self.assert_results(results, [{}])
def test_merge_chrom(self):
s1 = SNPs("tests/input/generic.csv")
df = s1.snps.append(
self.create_snp_df(
rsid=["rs100", "rs101", "rs102", "rs103"],
chrom=["Y", "Y", "Y", "Y"],
pos=[100, 101, 102, 103],
genotype=["A", np.nan, "A", "A"],
))
s1._snps = df.copy()
s2 = SNPs()
s2._build = 37
s2._snps = df.copy()
# set values for chrom that will be ignored (that would otherwise result in
# identification of discrepant SNPs or updating genotype)
s2._snps.loc["rs3", "pos"] = 1003 # discrepant position
s2._snps.loc["rs4", "genotype"] = "AA" # discrepant genotype
s2._snps.loc["rs5", "genotype"] = "AA"
# set values for chrom to be merged
s2._snps.loc["rs100", "genotype"] = "T" # discrepant genotype
s2._snps.loc["rs101", "genotype"] = "A"
s2._snps.loc["rs102", "pos"] = 1002 # discrepant position
# set expected values for merge result
df.loc["rs100",
"genotype"] = np.nan # discrepant genotype sets to np.nan
df.loc["rs101", "genotype"] = "A" # updates np.nan
results = s1.merge([s2], chrom="Y")
pd.testing.assert_frame_equal(s1.snps, df, check_exact=True)
self.assert_results(
results,
[{
"merged":
True,
"common_rsids":
pd.Index(["rs100", "rs101", "rs102", "rs103"], name="rsid"),
"discrepant_position_rsids":
pd.Index(["rs102"], name="rsid"),
"discrepant_genotype_rsids":
pd.Index(["rs100"], name="rsid"),
}],
)
self.assertEqual(len(s1.discrepant_merge_positions), 1)
self.assertEqual(len(s1.discrepant_merge_genotypes), 1)
def test_source_snps(self):
with tempfile.TemporaryDirectory() as tmpdir:
s = SNPs("tests/input/GRCh37.csv", output_dir=tmpdir)
self.assertEqual(s.source, "generic")
results = s.merge((SNPs("tests/input/23andme.txt"), ))
self.assertEqual(s.source, "generic, 23andMe")
self.assertListEqual(s._source, ["generic", "23andMe"])
dest = os.path.join(tmpdir, "generic__23andMe_GRCh37.txt")
self.assertEqual(s.save(), dest)
s = SNPs(dest)
self.assertEqual(s.source, "generic, 23andMe")
self.assertListEqual(s._source, ["generic", "23andMe"])
pd.testing.assert_frame_equal(s.snps, s.snps, check_exact=True)
self.assert_results(results, [{"merged": True}])
def test_source_snps(self):
s = SNPs("tests/input/GRCh37.csv")
self.assertEqual(s.source, "generic")
results = s.merge((SNPs("tests/input/23andme.txt"),))
self.assertEqual(s.source, "generic, 23andMe")
self.assertListEqual(s._source, ["generic", "23andMe"])
self.assertEqual(
os.path.relpath(s.save()), f"output{os.sep}generic__23andMe_GRCh37.txt"
)
s = SNPs("output/generic__23andMe_GRCh37.txt")
self.assertEqual(s.source, "generic, 23andMe")
self.assertListEqual(s._source, ["generic", "23andMe"])
pd.testing.assert_frame_equal(s.snps, s.snps, check_exact=True)
self.assert_results(results, [{"merged": True}])
def test_merge_remap_false(self):
s = SNPs("tests/input/NCBI36.csv")
results = s.merge([SNPs("tests/input/GRCh37.csv")], remap=False)
self.assertEqual(len(s.discrepant_merge_positions), 4)
pd.testing.assert_index_equal(
s.discrepant_merge_positions.index,
results[0]["discrepant_position_rsids"],
check_exact=True,
check_names=True,
)
self.assertEqual(len(s.discrepant_merge_genotypes), 1)
pd.testing.assert_index_equal(
s.discrepant_merge_genotypes.index,
results[0]["discrepant_genotype_rsids"],
check_exact=True,
check_names=True,
)
self.assertEqual(len(s.discrepant_merge_positions_genotypes), 4)
pd.testing.assert_index_equal(
s.discrepant_merge_positions_genotypes.index,
results[0]["discrepant_position_rsids"],
check_exact=True,
check_names=True,
)
expected = self.snps_NCBI36()
expected.loc[
"rs11928389",
"genotype"] = np.nan # discrepant genotype is set to null / NA
pd.testing.assert_frame_equal(s.snps, expected, check_exact=True)
self.assert_results(
results,
[{
"merged":
True,
"common_rsids":
pd.Index(
["rs3094315", "rs2500347", "rsIndelTest", "rs11928389"],
name="rsid",
),
"discrepant_position_rsids":
pd.Index(
["rs3094315", "rs2500347", "rsIndelTest", "rs11928389"],
name="rsid",
),
"discrepant_genotype_rsids":
pd.Index(["rs11928389"], name="rsid"),
}],
)
def f():
s = SNPs("tests/input/NCBI36.csv")
results = s.merge([SNPs("tests/input/GRCh37.csv")])
self.assertEqual(len(s.discrepant_merge_positions), 0)
self.assertEqual(len(s.discrepant_merge_genotypes), 0)
pd.testing.assert_frame_equal(s.snps, self.snps_NCBI36(), check_exact=True)
self.assert_results(
results,
[
{
"merged": True,
"common_rsids": pd.Index(
["rs3094315", "rs2500347", "rsIndelTest", "rs11928389"],
name="rsid",
),
}
],
)
def test_merge_unphased(self):
s1 = SNPs("tests/input/generic.csv")
s2 = SNPs("tests/input/generic.csv")
s1._phased = True
results = s1.merge([s2])
self.assertFalse(s1.phased)
pd.testing.assert_frame_equal(s1.snps, self.generic_snps(), check_exact=True)
self.assert_results(
results,
[
{
"merged": True,
"common_rsids": pd.Index(
["rs1", "rs2", "rs3", "rs4", "rs5", "rs6", "rs7", "rs8"],
name="rsid",
),
}
],
)
def test_merging_files_discrepant_snps(self):
with tempfile.TemporaryDirectory() as tmpdir:
dest1 = os.path.join(tmpdir, "discrepant_snps1.csv")
dest2 = os.path.join(tmpdir, "discrepant_snps2.csv")
df = pd.read_csv(
"tests/input/discrepant_snps.csv",
skiprows=1,
na_values="--",
names=[
"rsid",
"chrom",
"pos_file1",
"pos_file2",
"genotype_file1",
"genotype_file2",
"discrepant_position",
"discrepant_genotype",
"expected_position",
"expected_genotype",
],
index_col=0,
dtype={
"chrom": object,
"pos_file1": np.uint32,
"pos_file2": np.uint32,
"discrepant_position": bool,
"discrepant_genotype": bool,
"expected_position": np.uint32,
},
)
df1 = df[["chrom", "pos_file1", "genotype_file1"]]
df2 = df[["chrom", "pos_file2", "genotype_file2"]]
df1.to_csv(dest1,
na_rep="--",
header=["chromosome", "position", "genotype"])
df2.to_csv(dest2,
na_rep="--",
header=["chromosome", "position", "genotype"])
s = SNPs()
s.merge([SNPs(dest1), SNPs(dest2)])
expected = df[[
"chrom",
"discrepant_position",
"discrepant_genotype",
"expected_position",
"expected_genotype",
]]
expected = expected.rename(columns={
"expected_position": "pos",
"expected_genotype": "genotype"
})
expected_snps = SNPs()
expected_snps._snps = expected
expected_snps.sort()
expected = expected_snps.snps
pd.testing.assert_index_equal(
s.discrepant_merge_positions.index,
expected.loc[expected["discrepant_position"] == True].index,
)
pd.testing.assert_index_equal(
s.discrepant_merge_genotypes.index,
expected.loc[expected["discrepant_genotype"] == True].index,
)
pd.testing.assert_series_equal(s.snps["pos"], expected["pos"])
pd.testing.assert_series_equal(s.snps["genotype"],
expected["genotype"])