def helper_check_region_with_chromosome(self, chromosome):
start = 1
stop = 8
regions_file_path = create_single_region_gff_file(
self.dir_path, chromosome, start, stop)
bam_file_path = create_bam(self.dir_path, [chromosome], self.sequence)
liquidator = blb.RegionLiquidator(regions_file=regions_file_path,
output_directory=os.path.join(
self.dir_path, 'output'),
bam_file_path=bam_file_path)
with tables.open_file(liquidator.counts_file_path) as counts:
self.assertEqual(1, len(
counts.root.files)) # 1 since only a single bam file
self.assertEqual(
1,
counts.root.files[0]['length']) # 1 since only a single read
record = counts.root.region_counts[0]
self.assertEqual(start, record['start'])
self.assertEqual(stop, record['stop'])
self.assertEqual(
stop - start, record['count']
) # count represents how many base pair reads intersected
# the region
counts.root.files[0]['length']
factor = (1 / (stop - start)) * (1 / (1 / 10**6))
self.assertEqual(record['count'] * factor,
record['normalized_count'])
def test_empty_region_file(self):
empty_file_path = os.path.join(self.dir_path, 'empty.gff')
open(empty_file_path, 'w').close()
liquidator = blb.RegionLiquidator(regions_file=empty_file_path,
output_directory=os.path.join(
self.dir_path, 'output'),
bam_file_path=self.bam_file_path)
liquidator.batch(extension=0, sense='.')
def test_region_liquidation(self):
start = 1
stop = 8
region_name = 'region_f'
regions_file_path = create_single_region_gff_file(
self.dir_path,
self.chromosome,
start,
stop,
region_name=region_name)
liquidator = blb.RegionLiquidator(regions_file=regions_file_path,
output_directory=os.path.join(
self.dir_path, 'output'),
bam_file_path=self.bam_file_path)
liquidator.flatten()
matrix_path = os.path.join(self.dir_path, 'matrix.gff')
blb.write_bamToGff_matrix(matrix_path, liquidator.counts_file_path)
with tables.open_file(liquidator.counts_file_path) as counts:
self.assertEqual(1, len(
counts.root.files)) # 1 since only a single bam file
self.assertEqual(
1,
counts.root.files[0]['length']) # 1 since only a single read
self.assertEqual(1, len(
counts.root.region_counts)) # 1 since only a single region
record = counts.root.region_counts[0]
self.assertEqual(region_name, record['region_name'].decode())
self.assertEqual(start, record['start'])
self.assertEqual(stop, record['stop'])
self.assertEqual(
stop - start, record['count']
) # count represents how many base pair reads intersected
# the region
# todo: add normalization record checks
with open(matrix_path, 'r') as matrix_file:
matrix_lines = matrix_file.readlines()
self.assertEqual(2, len(matrix_lines))
header_cols = matrix_lines[0].split('\t')
self.assertEqual(3, len(header_cols))
self.assertEqual('GENE_ID', header_cols[0])
self.assertEqual('locusLine', header_cols[1])
self.assertEqual('bin_1_%s\n' % 'single.bam', header_cols[2])
data_cols = matrix_lines[1].split('\t')
self.assertEqual(3, len(data_cols))
self.assertEqual(region_name, data_cols[0])
self.assertEqual('chr1(.):1-8',
data_cols[1]) # todo: don't hardcode these values
self.assertEqual('1000000.0\n', data_cols[2])
def test_region_other_extension(self):
start = 1
stop = len(self.sequence)
regions_file_path = create_single_region_gff_file(
self.dir_path,
self.chromosome,
start,
stop,
file_name='single.txt')
region_liquidator = blb.RegionLiquidator(
regions_file=regions_file_path,
region_format='gff',
output_directory=os.path.join(self.dir_path, 'region_output'),
bam_file_path=self.bam_file_path)
def test_region_long_bam_file_name(self):
long_file_name = 'x' * 65 # more than Float64Col
long_file_path = os.path.join(self.dir_path, long_file_name)
shutil.copyfile(self.bam_file_path, long_file_path)
shutil.copyfile(self.bam_file_path + '.bai', long_file_path + '.bai')
start = 1
stop = len(self.sequence)
regions_file_path = create_single_region_gff_file(
self.dir_path, self.chromosome, start, stop)
region_liquidator = blb.RegionLiquidator(
regions_file=regions_file_path,
output_directory=os.path.join(self.dir_path, 'region_output'),
bam_file_path=long_file_path)
def test_region_with_wrong_chromosome(self):
start = len(self.sequence) + 10
stop = start + 10
regions_file_path = create_single_region_gff_file(
self.dir_path, self.chromosome + '0', start, stop)
liquidator = blb.RegionLiquidator(regions_file=regions_file_path,
output_directory=os.path.join(
self.dir_path, 'output'),
bam_file_path=self.bam_file_path)
with tables.open_file(liquidator.counts_file_path) as counts:
self.assertEqual(1, len(
counts.root.files)) # 1 since only a single bam file
self.assertEqual(
1,
counts.root.files[0]['length']) # 1 since only a single read
self.assertEqual(0, len(
counts.root.region_counts)) # no valid regions
def test_region_with_really_long_name(self):
start = 1
stop = 8
region_name = 'r' * 84
truncated_region_name = 'r' * 63
regions_file_path = create_single_region_gff_file(
self.dir_path,
self.chromosome,
start,
stop,
region_name=region_name)
liquidator = blb.RegionLiquidator(regions_file=regions_file_path,
output_directory=os.path.join(
self.dir_path, 'output'),
bam_file_path=self.bam_file_path)
liquidator.flatten()
matrix_path = os.path.join(self.dir_path, 'matrix.gff')
blb.write_bamToGff_matrix(matrix_path, liquidator.counts_file_path)
with tables.open_file(liquidator.counts_file_path) as counts:
self.assertEqual(1, len(
counts.root.files)) # 1 since only a single bam file
self.assertEqual(
1,
counts.root.files[0]['length']) # 1 since only a single read
self.assertEqual(1, len(
counts.root.region_counts)) # 1 since only a single region
record = counts.root.region_counts[0]
self.assertEqual(truncated_region_name,
record['region_name'].decode())
self.assertEqual(start, record['start'])
self.assertEqual(stop, record['stop'])
self.assertEqual(
stop - start, record['count']
) # count represents how many base pair reads intersected
def test_region_liquidation_with_optional_bed_columns(self):
start = 1
stop = 8
region_name = 'region_f'
score = 9.9
strand = "."
columns = [
"chr1",
str(start),
str(stop), region_name,
str(score), strand, "thickStart", "thickEnd", "itemRgb",
"blockCount", "blockSizes", "blockStarts"
]
def create_single_region_bed_file(dir_path, columns):
region_file_path = os.path.join(dir_path, "single.bed")
with open(region_file_path, 'w') as region_file:
first_column_written = False
for column in columns:
if first_column_written:
region_file.write('\t')
region_file.write('%s' % column)
first_column_written = True
region_file.write('\n')
return region_file_path
for number_of_columns in range(0, 9):
regions_file_path = create_single_region_bed_file(
self.dir_path, columns[:number_of_columns])
try:
liquidator = blb.RegionLiquidator(
regions_file=regions_file_path,
output_directory=os.path.join(self.dir_path, 'output'),
bam_file_path=self.bam_file_path)
except Exception:
self.assertLess(number_of_columns, 3)
continue
else:
self.assertGreaterEqual(number_of_columns, 3)
liquidator.flatten()
matrix_path = os.path.join(self.dir_path, 'matrix.gff')
blb.write_bamToGff_matrix(matrix_path, liquidator.counts_file_path)
with tables.open_file(liquidator.counts_file_path) as counts:
self.assertEqual(1, len(
counts.root.files)) # 1 since only a single bam file
self.assertEqual(1, counts.root.files[0]
['length']) # 1 since only a single read
self.assertEqual(1, len(
counts.root.region_counts)) # 1 since only a single region
record = counts.root.region_counts[0]
if number_of_columns >= 4:
self.assertEqual(region_name,
record['region_name'].decode())
else:
self.assertEqual("", record['region_name'].decode())
self.assertEqual(start, record['start'])
self.assertEqual(stop, record['stop'])
self.assertEqual(
stop - start, record['count']
) # count represents how many base pair reads intersected
# the region
with open(matrix_path, 'r') as matrix_file:
matrix_lines = matrix_file.readlines()
self.assertEqual(2, len(matrix_lines))
header_cols = matrix_lines[0].split('\t')
self.assertEqual(3, len(header_cols))
self.assertEqual('GENE_ID', header_cols[0])
self.assertEqual('locusLine', header_cols[1])
self.assertEqual('bin_1_%s\n' % 'single.bam', header_cols[2])
data_cols = matrix_lines[1].split('\t')
self.assertEqual(3, len(data_cols))
if number_of_columns >= 4:
self.assertEqual(region_name, data_cols[0])
else:
self.assertEqual("", data_cols[0])
self.assertEqual(
'chr1(.):1-8',
data_cols[1]) # todo: don't hardcode these values
self.assertEqual('1000000.0\n', data_cols[2])