def test_combiner_plan_round_trip_serialization():
sample_names = all_samples[:5]
paths = [os.path.join(resource('gvcfs'), '1kg_chr22', f'{s}.hg38.g.vcf.gz') for s in sample_names]
plan_path = new_temp_file(extension='json')
out_file = new_temp_file(extension='vds')
plan = new_combiner(gvcf_paths=paths,
output_path=out_file,
temp_path=Env.hc()._tmpdir,
save_path=plan_path,
reference_genome='GRCh38',
use_exome_default_intervals=True,
branch_factor=2,
batch_size=2)
plan.save()
plan_loaded = load_combiner(plan_path)
assert plan == plan_loaded
def vds_combiner_chr22(*paths):
with TemporaryDirectory() as tmpdir:
with TemporaryDirectory() as outpath:
parts = hl.eval([
hl.parse_locus_interval('chr22:start-end',
reference_genome='GRCh38')
])
from hail.vds.combiner import new_combiner
combiner = new_combiner(output_path=outpath,
intervals=parts,
temp_path=tmpdir,
gvcf_paths=paths,
reference_genome='GRCh38',
branch_factor=16,
target_records=10000000)
combiner.run()
def test_combiner_manual_filtration():
sample_names = all_samples[:2]
paths = [os.path.join(resource('gvcfs'), '1kg_chr22', f'{s}.hg38.g.vcf.gz') for s in sample_names]
out_file = new_temp_file(extension='vds')
plan = new_combiner(gvcf_paths=paths,
output_path=out_file,
temp_path=Env.hc()._tmpdir,
reference_genome='GRCh38',
use_exome_default_intervals=True,
gvcf_reference_entry_fields_to_keep=['GQ'],
gvcf_info_to_keep=['ExcessHet'],
force=True)
assert plan.gvcf_info_to_keep == {'ExcessHet'}
plan.run()
vds = hl.vds.read_vds(out_file)
assert list(vds.variant_data.gvcf_info) == ['ExcessHet']
assert list(vds.reference_data.entry) == ['END', 'GQ']