def extract(self, input, output, format='single'):
if not os.path.exists(output):
os.makedirs(output)
batch_name, batch_ext = os.path.splitext(input)
# packed single reads in tar archive
if batch_ext == '.tar':
if format in ['single', 'lazy']:
with tarfile.open(input) as fp_tar:
fp_tar.extractall(path=output)
else:
with tempfile.TemporaryDirectory(
prefix=self.tmp_prefix) as tmpdirname, tarfile.open(
input) as fp_tar:
fp_tar.extractall(path=tmpdirname)
f5files = [
os.path.join(dirpath, f)
for dirpath, _, files in os.walk(tmpdirname)
for f in files if f.endswith('.fast5')
]
output_bulk_file = os.path.join(
output,
os.path.basename(batch_name) + '.fast5')
single_to_multi_fast5.create_multi_read_file(
f5files, output_bulk_file)
# bulk fast5
elif batch_ext == '.fast5':
if format in ['bulk', 'lazy']:
shutil.copy(input, output)
else:
multi_to_single_fast5.convert_multi_to_single(
input, output, '')
# read IDs to be extracted
elif batch_ext == '.txt':
# load index and requested IDs
if not self.index_dict:
raise RuntimeError(
"[Error] Extraction of reads from IDs without index file provided."
)
with open(input, 'r') as fp:
batch_ids = [id.strip() for id in fp.read().split('\n') if id]
if format in ['single', 'lazy']:
self.__copy_reads_to__(batch_ids, output)
else:
with tempfile.TemporaryDirectory(
prefix=self.tmp_prefix) as tmpdirname:
self.__copy_reads_to__(batch_ids, tmpdirname)
f5files = [
os.path.join(dirpath, f)
for dirpath, _, files in os.walk(tmpdirname)
for f in files if f.endswith('.fast5')
]
output_bulk_file = os.path.join(
output,
os.path.basename(batch_name) + '.fast5')
single_to_multi_fast5.create_multi_read_file(
f5files, output_bulk_file)
else:
raise RuntimeError(
'[ERROR] Raw fast5 batch extension {} not supported.'.format(
batch_ext))
def check_file_type(myfile):
fobj = fast5_interface.get_fast5_file(os.path.join(root, name))
if fast5_interface.check_file_type(fobj) == "multi-read":
#convert file to single fast5
print("converting fast5 file****")
multi_to_single_fast5.convert_multi_to_single(os.path.join(root, name),
directory, "single")
def convert_fast5_type(directory):
#go through fast5 files and check if the files is multi or single fast5 file
#we need a single fast5 file
for root, dirs, files in os.walk(directory):
for name in files:
if name.endswith(".fast5"):
fobj = fast5_interface.get_fast5_file(os.path.join(root, name))
if fast5_interface.check_file_type(fobj) == "multi-read":
#convert file to single fast5
print("converting fast5 file****")
multi_to_single_fast5.convert_multi_to_single(
os.path.join(root, name), directory, "single")
def test_multi_to_single(self):
input_file = os.path.join(test_data, "multi_read", "batch_0.fast5")
with MultiFast5File(input_file, 'r') as f5:
read_count = len(f5.handle)
expected_files = sorted([
os.path.join(self.save_path, "{}", i + '.fast5')
for i in f5.get_read_ids()
])
subfolder = '0'
convert_multi_to_single(input_file, self.save_path, subfolder)
out_files = sorted(
get_fast5_file_list(self.save_path,
recursive=True,
follow_symlinks=True))
self.assertEqual(len(out_files), read_count)
self.assertEqual(out_files,
[f.format(subfolder) for f in expected_files])
def test_multi_to_single(self):
input_file = os.path.join(test_data, "multi_read", "batch_0.fast5")
with MultiFast5File(input_file, 'r') as f5:
read_count = len(f5.handle)
expected_files = sorted([
os.path.join(save_path, "{}", i + '.fast5')
for i in f5.get_read_ids()
])
subfolder = '0'
convert_multi_to_single(input_file, save_path, subfolder)
out_files = sorted(get_fast5_file_list(save_path, recursive=True))
self.assertEqual(len(out_files), read_count)
self.assertEqual(out_files,
[f.format(subfolder) for f in expected_files])
# Small batch size should be split across multiple folders
shutil.rmtree(save_path)
