def test_concordance(self):
dataset = get_dataset()
glob_conc, cols_conc, rows_conc = hl.concordance(dataset, dataset)
self.assertEqual(sum([sum(glob_conc[i]) for i in range(5)]), dataset.count_rows() * dataset.count_cols())
counts = dataset.aggregate_entries(hl.Struct(n_het=agg.filter(dataset.GT.is_het(), agg.count()),
n_hom_ref=agg.filter(dataset.GT.is_hom_ref(),
agg.count()),
n_hom_var=agg.filter(dataset.GT.is_hom_var(),
agg.count()),
nNoCall=agg.filter(hl.is_missing(dataset.GT),
agg.count())))
self.assertEqual(glob_conc[0][0], 0)
self.assertEqual(glob_conc[1][1], counts.nNoCall)
self.assertEqual(glob_conc[2][2], counts.n_hom_ref)
self.assertEqual(glob_conc[3][3], counts.n_het)
self.assertEqual(glob_conc[4][4], counts.n_hom_var)
[self.assertEqual(glob_conc[i][j], 0) for i in range(5) for j in range(5) if i != j]
self.assertTrue(cols_conc.all(hl.sum(hl.flatten(cols_conc.concordance)) == dataset.count_rows()))
self.assertTrue(rows_conc.all(hl.sum(hl.flatten(rows_conc.concordance)) == dataset.count_cols()))
cols_conc.write('/tmp/foo.kt', overwrite=True)
rows_conc.write('/tmp/foo.kt', overwrite=True)
def test_concordance(self):
dataset = get_dataset()
glob_conc, cols_conc, rows_conc = hl.concordance(dataset, dataset)
self.assertEqual(sum([sum(glob_conc[i]) for i in range(5)]), dataset.count_rows() * dataset.count_cols())
counts = dataset.aggregate_entries(hl.Struct(n_het=agg.filter(dataset.GT.is_het(), agg.count()),
n_hom_ref=agg.filter(dataset.GT.is_hom_ref(),
agg.count()),
n_hom_var=agg.filter(dataset.GT.is_hom_var(),
agg.count()),
nNoCall=agg.filter(hl.is_missing(dataset.GT),
agg.count())))
self.assertEqual(glob_conc[0][0], 0)
self.assertEqual(glob_conc[1][1], counts.nNoCall)
self.assertEqual(glob_conc[2][2], counts.n_hom_ref)
self.assertEqual(glob_conc[3][3], counts.n_het)
self.assertEqual(glob_conc[4][4], counts.n_hom_var)
[self.assertEqual(glob_conc[i][j], 0) for i in range(5) for j in range(5) if i != j]
self.assertTrue(cols_conc.all(hl.sum(hl.flatten(cols_conc.concordance)) == dataset.count_rows()))
self.assertTrue(rows_conc.all(hl.sum(hl.flatten(rows_conc.concordance)) == dataset.count_cols()))
cols_conc.write('/tmp/foo.kt', overwrite=True)
rows_conc.write('/tmp/foo.kt', overwrite=True)
def compute_concordance(mt: hl.MatrixTable, other_mt: hl.MatrixTable,
name: str) -> Tuple[hl.Table, hl.Table]:
# Filter to sites present in mt samples
mt = mt.filter_rows(hl.agg.any(mt.GT.is_non_ref()))
other_mt = other_mt.filter_rows(hl.agg.any(other_mt.GT.is_non_ref()))
summary, sample_concordance_ht, sites_concordance_ht = hl.concordance(
mt, other_mt)
logger.info(f'{name} concordance summary: %s', pformat(summary))
return sample_concordance_ht, sites_concordance_ht
def concordance_tables(full_vcf, downsample_dict, output, overwrite):
"""
runs concordance between full vcf and downsampled vcf
:param full_vcf:
:param downsample_dict:
:param output:
:param overwrite:
:return:
"""
global_conc, cols_conc, rows_conc = hl.concordance(full_vcf,
downsample_dict)
pprint(global_conc)
cols_conc.write(output + 'samples.ht', overwrite=overwrite)
rows_conc.write(output + 'variants.ht', overwrite=overwrite)
def test_concordance_no_values_doesnt_error(self):
dataset = get_dataset().filter_rows(False)
_, cols_conc, rows_conc = hl.concordance(dataset, dataset)
cols_conc._force_count()
rows_conc._force_count()
def test_concordance_n_discordant(self):
dataset = get_dataset()
_, cols_conc, rows_conc = hl.concordance(dataset, dataset)
assert cols_conc.aggregate(
hl.agg.count_where(cols_conc.n_discordant != 0)) == 0
rows1 = [
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '1',
'GT': hl.Call([0, 0])
}),
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '2',
'GT': hl.Call([0, 0])
}),
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '3',
'GT': hl.Call([1, 1])
}),
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '4',
'GT': hl.Call([1, 1])
}),
hl.Struct(
**{
'locus': hl.Locus('1', 101),
'alleles': ['A', 'T'],
's': '1',
'GT': hl.Call([1, 1])
}),
]
rows2 = [
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '1',
'GT': None
}),
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '2',
'GT': hl.Call([0, 1])
}),
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '3',
'GT': hl.Call([0, 1])
}),
hl.Struct(
**{
'locus': hl.Locus('1', 100),
'alleles': ['A', 'T'],
's': '4',
'GT': hl.Call([1, 1])
}),
]
def make_mt(rows):
ht = hl.Table.parallelize(
rows,
schema=
'struct{locus:locus<GRCh37>,alleles:array<str>,s:str,GT:call}')
return ht.to_matrix_table(row_key=['locus', 'alleles'],
col_key=['s'])
global_conc_2, cols_conc_2, rows_conc_2 = hl.concordance(
make_mt(rows1), make_mt(rows2))
assert cols_conc_2.collect() == [
hl.Struct(s='1',
concordance=[[0, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 1, 0, 0, 0], [0, 0, 0, 0, 0],
[1, 0, 0, 0, 0]],
n_discordant=0),
hl.Struct(s='2',
concordance=[[1, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 1, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 0, 0]],
n_discordant=1),
hl.Struct(s='3',
concordance=[[1, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 1, 0]],
n_discordant=1),
hl.Struct(s='4',
concordance=[[1, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 0, 1]],
n_discordant=0),
]
assert global_conc_2 == [[3, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 1, 0, 1, 0], [0, 0, 0, 0, 0],
[1, 0, 0, 1, 1]]
assert rows_conc_2.collect() == [
hl.Struct(locus=hl.Locus('1', 100),
alleles=['A', 'T'],
concordance=[[0, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 1, 0, 1, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 1, 1]],
n_discordant=2),
hl.Struct(locus=hl.Locus('1', 101),
alleles=['A', 'T'],
concordance=[[3, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[0, 0, 0, 0, 0], [0, 0, 0, 0, 0],
[1, 0, 0, 0, 0]],
n_discordant=0),
]
pipeline_ds.describe()
autosomes_interval_expressions = []
for chrom in range(1,23):
interval_expr = hl.parse_locus_interval('chr' + str(chrom), reference_genome=reference)
autosomes_interval_expressions.append(interval_expr)
giab_auts_ds = hl.filter_intervals(giab_ds, autosomes_interval_expressions)
pipeline_auts_ds = hl.filter_intervals(pipeline_ds, autosomes_interval_expressions)
giab_auts_ds = giab_auts_ds.annotate_entries(GT = hl.call(giab_auts_ds.GT[0], giab_auts_ds.GT[1], phased=False))
pipeline_auts_ds = pipeline_auts_ds.annotate_entries(GT = hl.call(pipeline_auts_ds.GT[0], pipeline_auts_ds.GT[1], phased=False))
# Run genotype concordance
global_conc, cols_conc, rows_conc = hl.concordance(giab_auts_ds, pipeline_auts_ds)
summary = global_conc
left_homref_right_homvar = summary[2][4]
left_het_right_missing = summary[3][1]
left_het_right_something_else = sum(summary[3][:]) - summary[3][3]
total_concordant = summary[2][2] + summary[3][3] + summary[4][4]
total_discordant = sum([sum(s[2:]) for s in summary[2:]]) - total_concordant
concordance = total_concordant/float(total_concordant + total_discordant)
now = datetime.datetime.utcnow()
results_bucket = "{}/validation/{}/validation-result-{}{}{}.txt".format(bucket_name, cohort_prefix, now.month, now.day, now.year)
