def test_calc_dp_means(self):
snps = VariationsH5(join(TEST_DATA_DIR, 'ril.hdf5'), mode='r')
means = calc_depth_mean_by_sample(snps)
means2 = calc_depth_mean_by_sample(snps, chunk_size=None)
assert means.shape[0] == 153
assert numpy.allclose(means, means2)
def test_calc_dp_means(self):
snps = VariationsH5(join(TEST_DATA_DIR, 'ril.hdf5'), mode='r')
means = calc_depth_mean_by_sample(snps)
means2 = calc_depth_mean_by_sample(snps, chunk_size=None)
assert means.shape[0] == 153
assert numpy.allclose(means, means2)
def test_filter_samples_by_pseudohet(self):
snps = VariationsH5(join(TEST_DATA_DIR, 'ril.hdf5'), mode='r')
sample_dp_means = calc_depth_mean_by_sample(snps)
flt = PseudoHetDuplicationFilter2(sample_dp_means=sample_dp_means,
max_high_dp_freq=0.01,
do_histogram=True,
report_selection=True)
res = flt(snps)
assert res[FLT_STATS]['tot'] == snps.num_variations
assert min(res[EDGES]) < 1
assert max(res[EDGES]) > 0
assert res[FLT_VARS].num_variations == res[FLT_STATS]['n_kept']
assert res[SELECTED_VARS].shape
# some samples
samples = snps.samples[:50]
sample_dp_means = sample_dp_means[:50]
flt = PseudoHetDuplicationFilter2(sample_dp_means=sample_dp_means,
max_high_dp_freq=0.01,
do_histogram=True,
samples=samples)
res = flt(snps)
assert res[FLT_STATS]['tot'] == snps.num_variations
assert min(res[EDGES]) < 1
assert max(res[EDGES]) > 0
assert res[FLT_VARS].num_variations == res[FLT_STATS]['n_kept']
def test_filter_samples_by_pseudohet(self):
snps = VariationsH5(join(TEST_DATA_DIR, 'ril.hdf5'), mode='r')
sample_dp_means = calc_depth_mean_by_sample(snps)
flt = PseudoHetDuplicationFilter2(sample_dp_means=sample_dp_means,
max_high_dp_freq=0.01,
do_histogram=True,
report_selection=True)
res = flt(snps)
assert res[FLT_STATS]['tot'] == snps.num_variations
assert min(res[EDGES]) < 1
assert max(res[EDGES]) > 0
assert res[FLT_VARS].num_variations == res[FLT_STATS]['n_kept']
assert res[SELECTED_VARS].shape
# some samples
samples = snps.samples[:50]
sample_dp_means = sample_dp_means[:50]
flt = PseudoHetDuplicationFilter2(sample_dp_means=sample_dp_means,
max_high_dp_freq=0.01,
do_histogram=True,
samples=samples)
res = flt(snps)
assert res[FLT_STATS]['tot'] == snps.num_variations
assert min(res[EDGES]) < 1
assert max(res[EDGES]) > 0
assert res[FLT_VARS].num_variations == res[FLT_STATS]['n_kept']