def count_exact_matches(input, output, reference, read_length, sample):
"""Match reads to reference exactly.
Takes the first <read-length> bases of each read and attempt to match it
exactly to the reference sequences. Computes the number of matches for each
reference.
"""
# load reference
seq_to_ref = OrderedDict()
with open(reference, "r") as ip:
for (ref_name, seq, _) in readfq(ip):
seq_to_ref[seq[:read_length]] = ref_name
num_reads = 0
num_matched = 0
counts = Counter()
with open_maybe_compressed(input, "r") as ip:
for (name, seq, _) in tqdm(readfq(ip)):
num_reads += 1
refname = seq_to_ref.get(seq[:read_length])
if refname is not None:
num_matched += 1
counts[refname] += 1
print(
"num_reads: {}\nnum_matched: {}\nfrac_matched: {}".format(
num_reads, num_matched, num_matched / num_reads
),
file=sys.stderr,
)
if not sample:
sample = pathlib.Path(input).stem
with open(output, "w") as op:
print(f"id\t{sample}", file=op)
for (_, refname) in seq_to_ref.items():
print(f"{refname}\t{counts[refname]}", file=op)
def split_fastq(input, output, chunk_size):
"""Split fastq files into smaller chunks."""
input_file = osp.abspath(input)
output_dir = osp.abspath(output)
os.makedirs(output_dir, mode=0o755)
# convenience functions
output_file = lambda i: pjoin(output_dir, "part.{0}.fastq".format(i))
with open_maybe_compressed(input_file, "r") as input_handle:
num_processed = 0
file_num = 1
for (name, seq, qual) in readfq(input_handle):
if num_processed == 0:
op = open_maybe_compressed(output_file(file_num), "w")
print(f"@{name}\n{seq}\n+\n{qual}", file=op)
num_processed += 1
if num_processed == chunk_size:
op.close()
num_processed = 0
file_num += 1
if not op.closed:
op.close()
def truncate_fasta(input, output, length):
"""Truncate each sequence of a fasta file."""
with open(input, "r") as ip, open(output, "w") as op:
for (n, s, q) in readfq(ip):
print(f">{n}\n{s[:length]}", file=op)