def test_coverages(self):
"""Test that the regions and calculated coverages are the same for both the bfx and pbpy summarize_coverage results.
"""
# Read the pbpy gff into a dictionary for comparison
pbpy_gff_reader = GffIO.GffReader(self.gff_path)
pbpy_gff_records = {}
for gff_record in pbpy_gff_reader:
record_key = (gff_record.seqid.split()[0], gff_record.start,
gff_record.end)
record_val = gff_record
pbpy_gff_records[record_key] = record_val
# Recapitulate the first few steps of summarize_coverage.main
ds_reader, readers = self._get_readers()
interval_lists = summarize_coverage.build_interval_lists(readers)
get_region_size_frozen = functools.partial(
summarize_coverage.get_region_size,
num_refs=len(interval_lists),
region_size=0,
num_regions=500,
force_num_regions=False)
for ref_group_id in sorted(interval_lists):
gff_generator = summarize_coverage.generate_gff_records(
interval_lists[ref_group_id], readers, ref_group_id,
get_region_size_frozen, {})
for bfx_gff_record in gff_generator:
bfx_key = (bfx_gff_record.seqid, bfx_gff_record.start,
bfx_gff_record.end)
self.assertIn(bfx_key, pbpy_gff_records)
pbpy_gff_record = pbpy_gff_records.pop(bfx_key)
self.assertEqual(pbpy_gff_record.cov, bfx_gff_record.cov)
self.assertEqual(pbpy_gff_record.gaps, bfx_gff_record.gaps)
pbpy_cov2 = [float(k) for k in pbpy_gff_record.cov2.split(',')]
bfx_cov2 = [float(k) for k in bfx_gff_record.cov2.split(',')]
for pair in zip(pbpy_cov2, bfx_cov2):
self.assertAlmostEqual(pair[0], pair[1])
self.assertEqual(pbpy_gff_record.source, bfx_gff_record.source)
self.assertEqual(pbpy_gff_record.type, bfx_gff_record.type)
self.assertEqual(pbpy_gff_record.score, bfx_gff_record.score)
self.assertEqual(pbpy_gff_record.strand, bfx_gff_record.strand)
self.assertEqual(pbpy_gff_record.phase, bfx_gff_record.phase)
if self.selected_reference is not None:
remaining_pbpy_records = {}
for record_key in pbpy_gff_records:
if record_key[0] == self.selected_reference:
remaining_pbpy_records[record_key] = pbpy_gff_records[
record_key]
else:
remaining_pbpy_records = pbpy_gff_records
self.assertEqual(len(remaining_pbpy_records), 0)
def test_metadata(self):
"""Test that the metadata lines match those from the pbpy version of summarize_coverage.
"""
ds_reader, bam_readers = self._get_readers()
bfx_metadata_lines = summarize_coverage.get_metadata_lines(
bam_readers, {})
pbpy_gff_reader = GffIO.GffReader(self.gff_path)
pbpy_metadata_lines = pbpy_gff_reader.headers
pbpy_i = 0
bfx_i = 0
# GffWriter handles this first line, not get_metadata_lines. So just
# make sure it's there in the pbpy file
self.assertEqual(pbpy_metadata_lines[0], "##gff-version 3")
pbpy_i += 1
self.assertEqual(len(pbpy_metadata_lines[1:]), len(bfx_metadata_lines))
# Check the date line
bfx_dateline = bfx_metadata_lines[bfx_i]
pbpy_dateline = pbpy_metadata_lines[pbpy_i]
self.assertEqual(
bfx_dateline.split(' ')[0],
pbpy_dateline.split(' ')[0])
pbpy_i += 1
bfx_i += 1
# Check the source line
bfx_sourceline = bfx_metadata_lines[bfx_i]
pbpy_sourceline = pbpy_metadata_lines[pbpy_i]
self.assertEqual(bfx_sourceline, pbpy_sourceline)
pbpy_i += 1
bfx_i += 1
# Check the command line
bfx_commandline = bfx_metadata_lines[bfx_i]
pbpy_commandline = pbpy_metadata_lines[pbpy_i]
self.assertEqual(
bfx_commandline.split(' ')[0],
pbpy_commandline.split(' ')[0])
pbpy_i += 1
bfx_i += 1
while pbpy_i < len(pbpy_metadata_lines):
self.assertEqual(pbpy_metadata_lines[pbpy_i],
bfx_metadata_lines[bfx_i])
pbpy_i += 1
bfx_i += 1