def trim(reads, find_start=None, find_end=None, second_time=False):
''' Wrapper that handles the logistics of trimming reads given functions
find_start and find_end that take a sequence and returns positions
that trimming should occur at.
'''
if find_start == None:
find_start = lambda seq: 0
if find_end == None:
find_end = len
for read in reads:
start = find_start(read.seq)
end = find_end(read.seq)
left_seq = read.seq[:start]
left_qual = fastq.sanitize_qual(read.qual[:start])
right_seq = read.seq[end:]
right_qual = fastq.sanitize_qual(read.qual[end:])
if second_time:
payload_annotation = PayloadAnnotation.from_identifier(read.name)
annotation = TrimmedTwiceAnnotation(
retrimmed_left_seq=left_seq,
retrimmed_left_qual=left_qual,
retrimmed_right_seq=right_seq,
retrimmed_right_qual=right_qual,
**payload_annotation)
else:
annotation = PayloadAnnotation(
original_name=read.name,
left_seq=left_seq,
left_qual=left_qual,
right_seq=right_seq,
right_qual=right_qual,
)
trimmed_read = fastq.Read(
annotation.identifier,
read.seq[start:end],
read.qual[start:end],
)
yield trimmed_read
def trim(reads, find_start=None, find_end=None, second_time=False):
''' Wrapper that handles the logistics of trimming reads given functions
find_start and find_end that take a sequence and returns positions
that trimming should occur at.
'''
if find_start == None:
find_start = lambda seq: 0
if find_end == None:
find_end = len
for read in reads:
start = find_start(read.seq)
end = find_end(read.seq)
left_seq = read.seq[:start]
left_qual = fastq.sanitize_qual(read.qual[:start])
right_seq = read.seq[end:]
right_qual = fastq.sanitize_qual(read.qual[end:])
if second_time:
payload_annotation = PayloadAnnotation.from_identifier(read.name)
annotation = TrimmedTwiceAnnotation(retrimmed_left_seq=left_seq,
retrimmed_left_qual=left_qual,
retrimmed_right_seq=right_seq,
retrimmed_right_qual=right_qual,
**payload_annotation)
else:
annotation = PayloadAnnotation(original_name=read.name,
left_seq=left_seq,
left_qual=left_qual,
right_seq=right_seq,
right_qual=right_qual,
)
trimmed_read = fastq.Read(annotation.identifier,
read.seq[start:end],
read.qual[start:end],
)
yield trimmed_read
def find_boundary_sequences(R1, R2, counters):
# Find which read in the read pair is from the reverse strand by looking for
# common_right_reverse.
# First try to find a unique position entirely contained within R1 or R2
# that is close to common_right_reverse.
# Failing this, find the longest of (the longest suffix of R1 or R2 that
# matches a prefix of common_right_reverse) or (the longest prefix of R1 or
# R2 that matches a suffix of common_right_reverse).
R1_contained, R1_prefix, R1_suffix = all_adapter_possibilites(
R1.seq, common_right_reverse)
R2_contained, R2_prefix, R2_suffix = all_adapter_possibilites(
R2.seq, common_right_reverse)
if len(R1_contained) + len(R2_contained) > 1:
# Only one of occurence of common_right_reverse should exist between R1
# and R2.
return None, None
elif len(R1_contained) + len(R2_contained) == 0:
possiblities = [
(len(common_right_reverse) - R1_prefix, 'R1_prefix'),
(len(common_right_reverse) - R2_prefix, 'R2_prefix'),
(len(common_right_reverse) - R1_suffix, 'R1_suffix'),
(len(common_right_reverse) - R2_suffix, 'R2_suffix'),
]
length, kind = max(possiblities)
if length > 5:
if 'R1' in kind:
reverse_read = R1
forward_read = R2
polyA_read = 'R2_forward'
polyT_read = 'R1_reverse'
elif 'R2' in kind:
reverse_read = R2
forward_read = R1
polyA_read = 'R1_forward'
polyT_read = 'R2_reverse'
if 'prefix' in kind:
common_right_reverse_start = len(reverse_read.seq) - length
elif 'suffix' in kind:
common_right_reverse_start = -length
else:
return None, None
elif len(R1_contained) == 1:
reverse_read = R1
forward_read = R2
polyA_read = 'R2_forward'
polyT_read = 'R1_reverse'
common_right_reverse_start = R1_contained.pop()
elif len(R2_contained) == 1:
reverse_read = R2
forward_read = R1
polyA_read = 'R1_forward'
polyT_read = 'R2_reverse'
common_right_reverse_start = R2_contained.pop()
# '*' means that there was no opportunity to see this id.
# 'X' means that there was an opportunity and it was neither A nor B.
right_id = '*'
left_id = '*'
five_payload_slice = slice(None, max(0, common_right_reverse_start))
five_payload_seq = utilities.reverse_complement(
reverse_read.seq[five_payload_slice])
five_payload_qual = reverse_read.qual[five_payload_slice][::-1]
current_p = common_right_reverse_start + len(common_right_reverse)
if current_p < len(reverse_read.seq) - after_right_length:
right_id_seq = reverse_read.seq[current_p:current_p +
after_right_length]
for key, prefix in after_right_prefix.items():
if right_id_seq == prefix:
right_id = key
if right_id == '*':
right_id = 'X'
counters['right_ids'][right_id_seq] += 1
if right_id != 'X':
current_p += len(after_right[right_id])
if current_p < len(reverse_read.seq) - 4:
left_id_seq = reverse_read.seq[current_p:current_p + 4]
for key, sequence in after_left.items():
if left_id_seq == sequence:
left_id = key
if left_id == '*':
left_id = 'X'
counters['left_ids'][left_id_seq] += 1
polyA_start, polyA_length = find_polyA_cython.find_polyA(
forward_read.seq, 15)
polyA_slice = slice(polyA_start, polyA_start + polyA_length)
polyA_seq = forward_read.seq[polyA_slice]
polyA_qual = fastq.sanitize_qual(forward_read.qual[polyA_slice])
three_payload_slice = slice(None, polyA_start)
three_payload_seq = forward_read.seq[three_payload_slice]
three_payload_qual = forward_read.qual[three_payload_slice]
common_name, _ = R1.name.rsplit(':', 1)
control_ids_string = '{0}-{1}'.format(left_id, right_id)
five_annotation = trim.PayloadAnnotation(
original_name=common_name,
left_seq=control_ids_string,
left_qual='',
right_seq='',
right_qual='',
)
three_annotation = trim.PayloadAnnotation(
original_name=common_name,
left_seq=control_ids_string,
left_qual='',
right_seq=polyA_seq,
right_qual=polyA_qual,
)
five_payload_read = fastq.Read(five_annotation.identifier,
five_payload_seq, five_payload_qual)
three_payload_read = fastq.Read(three_annotation.identifier,
three_payload_seq, three_payload_qual)
counters['positions'][polyT_read][max(0, common_right_reverse_start)] += 1
counters['positions'][polyA_read][polyA_start] += 1
counters['joint_lengths'][max(0, common_right_reverse_start),
polyA_start] += 1
counters['polyA_lengths'][polyA_length] += 1
counters['control_ids'][control_ids_string] += 1
if polyA_length < 13:
return None, None
return five_payload_read, three_payload_read
def find_boundary_sequences(R1, R2, counters):
# Find which read in the read pair is from the reverse strand by looking for
# common_right_reverse.
# First try to find a unique position entirely contained within R1 or R2
# that is close to common_right_reverse.
# Failing this, find the longest of (the longest suffix of R1 or R2 that
# matches a prefix of common_right_reverse) or (the longest prefix of R1 or
# R2 that matches a suffix of common_right_reverse).
R1_contained, R1_prefix, R1_suffix = all_adapter_possibilites(R1.seq, common_right_reverse)
R2_contained, R2_prefix, R2_suffix = all_adapter_possibilites(R2.seq, common_right_reverse)
if len(R1_contained) + len(R2_contained) > 1:
# Only one of occurence of common_right_reverse should exist between R1
# and R2.
return None, None
elif len(R1_contained) + len(R2_contained) == 0:
possiblities = [(len(common_right_reverse) - R1_prefix, 'R1_prefix'),
(len(common_right_reverse) - R2_prefix, 'R2_prefix'),
(len(common_right_reverse) - R1_suffix, 'R1_suffix'),
(len(common_right_reverse) - R2_suffix, 'R2_suffix'),
]
length, kind = max(possiblities)
if length > 5:
if 'R1' in kind:
reverse_read = R1
forward_read = R2
polyA_read = 'R2_forward'
polyT_read = 'R1_reverse'
elif 'R2' in kind:
reverse_read = R2
forward_read = R1
polyA_read = 'R1_forward'
polyT_read = 'R2_reverse'
if 'prefix' in kind:
common_right_reverse_start = len(reverse_read.seq) - length
elif 'suffix' in kind:
common_right_reverse_start = -length
else:
return None, None
elif len(R1_contained) == 1:
reverse_read = R1
forward_read = R2
polyA_read = 'R2_forward'
polyT_read = 'R1_reverse'
common_right_reverse_start = R1_contained.pop()
elif len(R2_contained) == 1:
reverse_read = R2
forward_read = R1
polyA_read = 'R1_forward'
polyT_read = 'R2_reverse'
common_right_reverse_start = R2_contained.pop()
# '*' means that there was no opportunity to see this id.
# 'X' means that there was an opportunity and it was neither A nor B.
right_id = '*'
left_id = '*'
five_payload_slice = slice(None, max(0, common_right_reverse_start))
five_payload_seq = utilities.reverse_complement(reverse_read.seq[five_payload_slice])
five_payload_qual = reverse_read.qual[five_payload_slice][::-1]
current_p = common_right_reverse_start + len(common_right_reverse)
if current_p < len(reverse_read.seq) - after_right_length:
right_id_seq = reverse_read.seq[current_p:current_p + after_right_length]
for key, prefix in after_right_prefix.items():
if right_id_seq == prefix:
right_id = key
if right_id == '*':
right_id = 'X'
counters['right_ids'][right_id_seq] += 1
if right_id != 'X':
current_p += len(after_right[right_id])
if current_p < len(reverse_read.seq) - 4:
left_id_seq = reverse_read.seq[current_p:current_p + 4]
for key, sequence in after_left.items():
if left_id_seq == sequence:
left_id = key
if left_id == '*':
left_id = 'X'
counters['left_ids'][left_id_seq] += 1
polyA_start, polyA_length = find_polyA_cython.find_polyA(forward_read.seq, 15)
polyA_slice = slice(polyA_start, polyA_start + polyA_length)
polyA_seq = forward_read.seq[polyA_slice]
polyA_qual = fastq.sanitize_qual(forward_read.qual[polyA_slice])
three_payload_slice = slice(None, polyA_start)
three_payload_seq = forward_read.seq[three_payload_slice]
three_payload_qual = forward_read.qual[three_payload_slice]
common_name, _ = R1.name.rsplit(':', 1)
control_ids_string = '{0}-{1}'.format(left_id, right_id)
five_annotation = trim.PayloadAnnotation(original_name=common_name,
left_seq=control_ids_string,
left_qual='',
right_seq='',
right_qual='',
)
three_annotation = trim.PayloadAnnotation(original_name=common_name,
left_seq=control_ids_string,
left_qual='',
right_seq=polyA_seq,
right_qual=polyA_qual,
)
five_payload_read = fastq.Read(five_annotation.identifier, five_payload_seq, five_payload_qual)
three_payload_read = fastq.Read(three_annotation.identifier, three_payload_seq, three_payload_qual)
counters['positions'][polyT_read][max(0, common_right_reverse_start)] += 1
counters['positions'][polyA_read][polyA_start] += 1
counters['joint_lengths'][max(0, common_right_reverse_start), polyA_start] += 1
counters['polyA_lengths'][polyA_length] += 1
counters['control_ids'][control_ids_string] += 1
if polyA_length < 13:
return None, None
return five_payload_read, three_payload_read
