def load_reads(seq_file, label=None, max_len=100):
read_list = []
seq_format = get_seq_format(seq_file)
_open = partial(gzip.open, mode='rt') if seq_format.endswith(
"gz") else open
seq_type = "fasta" if seq_format.startswith("fa") else "fastq"
if label is None:
with _open(seq_file) as fh:
for record in seq_parser(fh, seq_type):
read_list.append(record)
else:
with _open(seq_file) as fh:
for record in seq_parser(fh, seq_type):
seq = record[1]
read, rc_read = get_read_rc_with_maxlen(seq, max_len=100)
read_list.extend([(label, read), (label, rc_read)])
return read_list
def load_seqs(seq_file, label, step_size):
seq_step_list = []
seq_format = get_seq_format(seq_file)
_open = partial(gzip.open, mode='rt') if seq_format.endswith(
"gz") else open
seq_type = "fasta" if seq_format.startswith("fa") else "fastq"
with _open(seq_file) as fh:
for record in seq_parser(fh, seq_type):
seq = record[1]
seq_step_list.extend([(label, seq, step_size),
(label, str(Seq(seq).reverse_complement()), step_size)])
return seq_step_list
def get_seq_chunks(seq_file, chunk_size=1048576):
seq_format = get_seq_format(seq_file)
_open = partial(gzip.open, mode='rt') if seq_format.endswith(
"gz") else open
seq_type = "fasta" if seq_format.startswith("fa") else "fastq"
with _open(seq_file) as fh:
seq_iterator = seq_parser(fh, seq_type)
while True:
seqs_chunk = list(islice(seq_iterator, chunk_size))
if seqs_chunk:
yield seqs_chunk
else:
break
def all_seqs_x(seq_file, min_seq_length):
dataset = []
seq_format = get_seq_format(seq_file)
_open = partial(gzip.open,
mode='rt') if seq_format.endswith("gz") else open
seq_type = "fasta" if seq_format.startswith("fa") else "fastq"
with _open(seq_file) as fh:
# for record in SeqIO.parse(fh, seq_type): # parse_seq_file(seq_file):
# seq = str(record.seq).upper()
for record in seq_parser(fh, seq_type):
features = seq_to_feature(record[1], min_seq_length)
try:
dataset.append(features)
except NameError as e:
print(NameError("Can not concatenate the np array", e))
return dataset
def output_seq(self):
norrna_count = 0
seq_format = get_seq_format(self.input[0])
seq_type = "fasta" if seq_format.startswith("fa") else "fastq"
_open = partial(gzip.open, mode='rt') if seq_format.endswith("gz") else open
self.logger.info('Writing output non-rRNA sequences into file: {}...'.format(", ".join(self.output)))
if len(self.pred_labels) == 2:
if self.rrna is not None:
self.logger.info('Writing output rRNA sequences into file: {}...'.format(", ".join(self.rrna)))
rrna1_fh = open(self.rrna[0], "w")
rrna2_fh = open(self.rrna[1], "w")
with open(self.output[0], "w") as out1_fh, open(self.output[1], "w") as out2_fh:
with _open(self.input[0]) as r1_fh, _open(self.input[1]) as r2_fh:
for idx, (r1, r2) in enumerate(tqdm(zip(seq_parser(r1_fh, seq_type), seq_parser(r2_fh, seq_type)))):
if self.pred_labels[0][idx] == self.pred_labels[1][idx] == 0:
# seq_r2_header = record.id.replace("/1", "/2")
out1_fh.write('\n'.join(r1) + '\n')
out2_fh.write('\n'.join(r2) + '\n')
# SeqIO.write(r1, out1_fh, seq_type)
# SeqIO.write(r2, out2_fh, seq_type)
norrna_count += 1
elif self.pred_labels[0][idx] == self.pred_labels[1][idx] == 1:
if self.rrna is not None:
# seq_r2_header = record.id.replace("/1", "/2")
rrna1_fh.write('\n'.join(r1) + '\n')
rrna2_fh.write('\n'.join(r2) + '\n')
# SeqIO.write(r1, rrna1_fh, seq_type)
# SeqIO.write(r2, rrna2_fh, seq_type)
else:
if self.args.ensure == "rrna":
# seq_r2_header = record.id.replace("/1", "/2")
out1_fh.write('\n'.join(r1) + '\n')
out2_fh.write('\n'.join(r2) + '\n')
# SeqIO.write(r1, out1_fh, seq_type)
# SeqIO.write(r2, out2_fh, seq_type)
norrna_count += 1
elif self.args.ensure == "norrna":
if self.rrna is not None:
# seq_r2_header = record.id.replace("/1", "/2")
rrna1_fh.write('\n'.join(r1) + '\n')
rrna2_fh.write('\n'.join(r2) + '\n')
# SeqIO.write(r1, rrna1_fh, seq_type)
# SeqIO.write(r2, rrna2_fh, seq_type)
else:
continue
if self.rrna is not None:
rrna1_fh.close()
rrna2_fh.close()
else:
if self.rrna is not None:
self.logger.info('Writing output rENA sequences into file: {}...'.format(", ".join(self.rrna)))
rrna_fh = open(self.rrna[0], "w")
with open(self.output[0], "w") as out_fh:
with _open(self.input[0]) as read_fh:
for idx, record in enumerate(tqdm(seq_parser(read_fh, seq_type))):
if self.pred_labels[0][idx] == 0:
out_fh.write('\n'.join(record) + '\n')
# SeqIO.write(record, out_fh, seq_type)
norrna_count += 1
else:
if self.rrna is not None:
rrna_fh.write('\n'.join(record) + '\n')
# SeqIO.write(record, rrna_fh, seq_type)
if self.rrna is not None:
rrna_fh.close()
self.logger.info('Finished writing {} non-rRNA sequences!'.format(norrna_count))