def trim_polyA_ends(fastq_filename,
output_dir,
compressed=False,
min_polyA_len=3,
min_read_len=22):
"""
Trim polyA ends from reads.
"""
print "Trimming polyA trails from: %s" % (fastq_filename)
# Strip the trailing extension
output_basename = \
".".join(os.path.basename(fastq_filename).split(".")[0:-1])
output_basename = "%s.trimmed_polyA.fastq.gz" % (output_basename)
output_filename = os.path.join(output_dir, output_basename)
utils.make_dir(output_dir)
if os.path.isfile(output_filename):
print "SKIPPING: %s already exists!" % (output_filename)
return output_filename
print " - Outputting trimmed sequences to: %s" % (output_filename)
input_file = fastq_utils.read_open_fastq(fastq_filename)
output_file = fastq_utils.write_open_fastq(output_filename)
t1 = time.time()
for line in fastq_utils.read_fastq(input_file):
header, seq, header2, qual = line
if seq.endswith("A"):
# Skip sequences that do not end with at least N
# many As
if seq[-min_polyA_len:] != ("A" * min_polyA_len):
continue
# Get sequence stripped of contiguous strech of polyAs
stripped_seq = rstrip_stretch(seq, "A")
if len(stripped_seq) < min_read_len:
# Skip altogether reads that are shorter than
# the required length after trimming
continue
# Strip the quality scores to match trimmed sequence
new_qual = qual[0:len(stripped_seq)]
new_rec = (header, stripped_seq, header2, new_qual)
# Write the record with trimmed sequence back out to file
fastq_utils.write_fastq(output_file, new_rec)
t2 = time.time()
print "Trimming took %.2f mins." % ((t2 - t1) / 60.)
output_file.close()
return output_filename
def trim_polyA_ends(fastq_filename,
output_dir,
compressed=False,
min_polyA_len=3,
min_read_len=22):
"""
Trim polyA ends from reads.
"""
print "Trimming polyA trails from: %s" %(fastq_filename)
# Strip the trailing extension
output_basename = \
".".join(os.path.basename(fastq_filename).split(".")[0:-1])
output_basename = "%s.trimmed_polyA.fastq.gz" %(output_basename)
output_filename = os.path.join(output_dir, output_basename)
utils.make_dir(output_dir)
if os.path.isfile(output_filename):
print "SKIPPING: %s already exists!" %(output_filename)
return output_filename
print " - Outputting trimmed sequences to: %s" %(output_filename)
input_file = fastq_utils.read_open_fastq(fastq_filename)
output_file = fastq_utils.write_open_fastq(output_filename)
t1 = time.time()
for line in fastq_utils.read_fastq(input_file):
header, seq, header2, qual = line
if seq.endswith("A"):
# Skip sequences that do not end with at least N
# many As
if seq[-min_polyA_len:] != ("A" * min_polyA_len):
continue
# Get sequence stripped of contiguous strech of polyAs
stripped_seq = rstrip_stretch(seq, "A")
if len(stripped_seq) < min_read_len:
# Skip altogether reads that are shorter than
# the required length after trimming
continue
# Strip the quality scores to match trimmed sequence
new_qual = qual[0:len(stripped_seq)]
new_rec = (header, stripped_seq, header2, new_qual)
# Write the record with trimmed sequence back out to file
fastq_utils.write_fastq(output_file, new_rec)
t2 = time.time()
print "Trimming took %.2f mins." %((t2 - t1)/60.)
output_file.close()
return output_filename
def get_fastx_entries(fastx_filename,
fasta=False,
fastq=False):
"""
Get entries of FASTQ/FASTA file.
if fasta=True, read file as fasta regardless of extension.
if fastq=True, read file as fastq regardless of extension
"""
entries = []
fastx_type = get_fastx_type(fastx_filename)
if (fastx_type == "fasta") or fasta:
# It's a FASTA file
entries = fasta_utils.read_fasta(fastx_filename)
elif (fastx_type == "fastq") or fastq:
# It's a FASTQ file
entries = fastq_utils.read_fastq(fastx_filename)
return entries
def get_seq_cycle_profile(self, fastq_filename,
first_n_seqs=None):#sample):
"""
Compute the average 'N' bases (unable to sequence)
as a function of the position of the read.
"""
fastq_file = fastq_utils.read_open_fastq(fastq_filename)
fastq_entries = fastq_utils.read_fastq(fastq_file)
# Mapping from position in read to number of Ns
num_n_bases = defaultdict(int)
# Mapping from position in read to total number of
# reads in that position
num_reads = defaultdict(int)
num_entries = 0
print "Computing sequence cycle profile for: %s" %(fastq_filename)
if first_n_seqs != None:
print "Looking at first %d sequences only" %(first_n_seqs)
for entry in fastq_entries:
if first_n_seqs != None:
# Stop at requested number of entries if asked to
if num_entries >= first_n_seqs:
break
header1, seq, header2, qual = entry
seq_len = len(seq)
for n in range(seq_len):
if seq[n] == "N":
# Record occurrences of N
num_n_bases[n] += 1
num_reads[n] += 1
num_entries += 1
# Compute percentage of N along each position
percent_n = []
for base_pos in range(max(num_reads.keys())):
curr_percent_n = float(num_n_bases[base_pos]) / num_reads[base_pos]
percent_n.append(curr_percent_n)
return percent_n
