def _tabular_msa_to_phylip(obj, fh):
sequence_count = obj.shape.sequence
if sequence_count < 1:
raise PhylipFormatError(
"TabularMSA can only be written in PHYLIP format if there is at "
"least one sequence in the alignment.")
sequence_length = obj.shape.position
if sequence_length < 1:
raise PhylipFormatError(
"TabularMSA can only be written in PHYLIP format if there is at "
"least one position in the alignment.")
chunk_size = 10
labels = [str(label) for label in obj.index]
for label in labels:
if len(label) > chunk_size:
raise PhylipFormatError(
"``TabularMSA`` can only be written in PHYLIP format if all "
"sequence index labels have %d or fewer characters. Found "
"sequence with index label '%s' that exceeds this limit. Use "
"``TabularMSA.reassign_index`` to assign shorter index labels."
% (chunk_size, label))
fh.write('{0:d} {1:d}\n'.format(sequence_count, sequence_length))
fmt = '{0:%d}{1}\n' % chunk_size
for label, seq in zip(labels, obj):
chunked_seq = chunk_str(str(seq), chunk_size, ' ')
fh.write(fmt.format(label, chunked_seq))
def _alignment_to_phylip(obj, fh):
if obj.is_empty():
raise PhylipFormatError(
"Alignment can only be written in PHYLIP format if there is at "
"least one sequence in the alignment.")
sequence_length = obj.sequence_length()
if sequence_length == 0:
raise PhylipFormatError(
"Alignment can only be written in PHYLIP format if there is at "
"least one position in the alignment.")
chunk_size = 10
for id_ in obj.ids():
if len(id_) > chunk_size:
raise PhylipFormatError(
"Alignment can only be written in PHYLIP format if all "
"sequence IDs have %d or fewer characters. Found sequence "
"with ID '%s' that exceeds this limit. Use "
"Alignment.update_ids to assign shorter IDs." %
(chunk_size, id_))
sequence_count = obj.sequence_count()
fh.write('{0:d} {1:d}\n'.format(sequence_count, sequence_length))
fmt = '{0:%d}{1}\n' % chunk_size
for seq in obj:
chunked_seq = chunk_str(str(seq), chunk_size, ' ')
fh.write(fmt.format(seq.metadata['id'], chunked_seq))
def _alignment_to_phylip(obj, fh):
if obj.is_empty():
raise PhylipFormatError(
"Alignment can only be written in PHYLIP format if there is at "
"least one sequence in the alignment.")
sequence_length = obj.sequence_length()
if sequence_length == 0:
raise PhylipFormatError(
"Alignment can only be written in PHYLIP format if there is at "
"least one position in the alignment.")
chunk_size = 10
for id_ in obj.ids():
if len(id_) > chunk_size:
raise PhylipFormatError(
"Alignment can only be written in PHYLIP format if all "
"sequence IDs have %d or fewer characters. Found sequence "
"with ID '%s' that exceeds this limit. Use "
"Alignment.update_ids to assign shorter IDs." %
(chunk_size, id_))
sequence_count = obj.sequence_count()
fh.write('{0:d} {1:d}\n'.format(sequence_count, sequence_length))
fmt = '{0:%d}{1}\n' % chunk_size
for seq in obj:
chunked_seq = chunk_str(str(seq), chunk_size, ' ')
fh.write(fmt.format(seq.metadata['id'], chunked_seq))
def _generator_to_fasta(obj, fh, qual=FileSentinel,
id_whitespace_replacement='_',
description_newline_replacement=' ', max_width=None,
lowercase=None):
if max_width is not None:
if max_width < 1:
raise ValueError(
"Maximum line width must be greater than zero (max_width=%d)."
% max_width)
if qual is not None:
# define text wrapper for splitting quality scores here for
# efficiency. textwrap docs recommend reusing a TextWrapper
# instance when it is used many times. configure text wrapper to
# never break "words" (i.e., integer quality scores) across lines
qual_wrapper = textwrap.TextWrapper(
width=max_width, break_long_words=False,
break_on_hyphens=False)
formatted_records = _format_fasta_like_records(
obj, id_whitespace_replacement, description_newline_replacement,
qual is not None, lowercase)
for header, seq_str, qual_scores in formatted_records:
if max_width is not None:
seq_str = chunk_str(seq_str, max_width, '\n')
fh.write('>%s\n%s\n' % (header, seq_str))
if qual is not None:
qual_str = ' '.join(np.asarray(qual_scores, dtype=np.str))
if max_width is not None:
qual_str = qual_wrapper.fill(qual_str)
qual.write('>%s\n%s\n' % (header, qual_str))
def _tabular_msa_to_phylip(obj, fh):
sequence_count = obj.shape.sequence
if sequence_count < 1:
raise PhylipFormatError(
"TabularMSA can only be written in PHYLIP format if there is at "
"least one sequence in the alignment.")
sequence_length = obj.shape.position
if sequence_length < 1:
raise PhylipFormatError(
"TabularMSA can only be written in PHYLIP format if there is at "
"least one position in the alignment.")
chunk_size = 10
labels = [str(label) for label in obj.index]
for label in labels:
if len(label) > chunk_size:
raise PhylipFormatError(
"``TabularMSA`` can only be written in PHYLIP format if all "
"sequence index labels have %d or fewer characters. Found "
"sequence with index label '%s' that exceeds this limit. Use "
"``TabularMSA.reassign_index`` to assign shorter index labels."
% (chunk_size, label))
fh.write('{0:d} {1:d}\n'.format(sequence_count, sequence_length))
fmt = '{0:%d}{1}\n' % chunk_size
for label, seq in zip(labels, obj):
chunked_seq = chunk_str(str(seq), chunk_size, ' ')
fh.write(fmt.format(label, chunked_seq))
def _generator_to_fasta(obj,
fh,
qual=FileSentinel,
id_whitespace_replacement='_',
description_newline_replacement=' ',
max_width=None,
lowercase=None):
if max_width is not None:
if max_width < 1:
raise ValueError(
"Maximum line width must be greater than zero (max_width=%d)."
% max_width)
if qual is not None:
# define text wrapper for splitting quality scores here for
# efficiency. textwrap docs recommend reusing a TextWrapper
# instance when it is used many times. configure text wrapper to
# never break "words" (i.e., integer quality scores) across lines
qual_wrapper = textwrap.TextWrapper(width=max_width,
break_long_words=False,
break_on_hyphens=False)
formatted_records = _format_fasta_like_records(
obj, id_whitespace_replacement, description_newline_replacement, qual
is not None, lowercase)
for header, seq_str, qual_scores in formatted_records:
if max_width is not None:
seq_str = chunk_str(seq_str, max_width, '\n')
fh.write('>%s\n%s\n' % (header, seq_str))
if qual is not None:
qual_str = ' '.join(np.asarray(qual_scores, dtype=np.str))
if max_width is not None:
qual_str = qual_wrapper.fill(qual_str)
qual.write('>%s\n%s\n' % (header, qual_str))
def _format_chunked_seq(self, line_idxs, num_chars, column_width):
"""Format specified lines of chunked sequence data."""
lines = []
for line_idx in line_idxs:
seq_idx = line_idx * num_chars
chars = str(self._obj[seq_idx:seq_idx + num_chars])
chunked_chars = chunk_str(chars, self._chunk_size, ' ')
lines.append(('%d' % seq_idx).ljust(column_width) + chunked_chars)
return lines
def _format_chunked_seq(self, line_idxs, num_chars, column_width):
"""Format specified lines of chunked sequence data."""
lines = []
for line_idx in line_idxs:
seq_idx = line_idx * num_chars
chars = str(self._obj[seq_idx:seq_idx+num_chars])
chunked_chars = chunk_str(chars, self._chunk_size, ' ')
lines.append(('%d' % seq_idx).ljust(column_width) + chunked_chars)
return lines
def test_even_split(self):
self.assertEqual(chunk_str('abcdef', 6, ' '), 'abcdef')
self.assertEqual(chunk_str('abcdef', 3, ' '), 'abc def')
self.assertEqual(chunk_str('abcdef', 2, ' '), 'ab cd ef')
self.assertEqual(chunk_str('abcdef', 1, ' '), 'a b c d e f')
self.assertEqual(chunk_str('a', 1, ' '), 'a')
self.assertEqual(chunk_str('abcdef', 2, ''), 'abcdef')
def test_even_split(self):
self.assertEqual(chunk_str('abcdef', 6, ' '), 'abcdef')
self.assertEqual(chunk_str('abcdef', 3, ' '), 'abc def')
self.assertEqual(chunk_str('abcdef', 2, ' '), 'ab cd ef')
self.assertEqual(chunk_str('abcdef', 1, ' '), 'a b c d e f')
self.assertEqual(chunk_str('a', 1, ' '), 'a')
self.assertEqual(chunk_str('abcdef', 2, ''), 'abcdef')
def _serialize_origin(seq, indent=9):
'''Serialize seq to ORIGIN.
Parameters
----------
seq : str
'''
n = 1
line_size = 60
frag_size = 10
for i in range(0, len(seq), line_size):
line = seq[i:i+line_size]
s = '{n:>{indent}} {s}\n'.format(
n=n, indent=indent, s=chunk_str(line, frag_size, ' '))
if n == 1:
s = 'ORIGIN\n' + s
n = n + line_size
yield s
def _serialize_origin(seq, indent=9):
'''Serialize seq to ORIGIN.
Parameters
----------
seq : str
'''
n = 1
line_size = 60
frag_size = 10
for i in range(0, len(seq), line_size):
line = seq[i:i+line_size]
s = '{n:>{indent}} {s}\n'.format(
n=n, indent=indent, s=chunk_str(line, frag_size, ' '))
if n == 1:
s = 'ORIGIN\n' + s
n = n + line_size
yield s
def test_invalid_n(self):
with six.assertRaisesRegex(self, ValueError, 'n=0'):
chunk_str('abcdef', 0, ' ')
with six.assertRaisesRegex(self, ValueError, 'n=-42'):
chunk_str('abcdef', -42, ' ')
def test_uneven_split(self):
self.assertEqual(chunk_str('abcdef', 5, '|'), 'abcde|f')
self.assertEqual(chunk_str('abcdef', 4, '|'), 'abcd|ef')
self.assertEqual(chunk_str('abcdefg', 3, ' - '), 'abc - def - g')
def test_no_split(self):
self.assertEqual(chunk_str('', 2, '\n'), '')
self.assertEqual(chunk_str('a', 100, '\n'), 'a')
self.assertEqual(chunk_str('abcdef', 42, '|'), 'abcdef')
def test_invalid_n(self):
with six.assertRaisesRegex(self, ValueError, 'n=0'):
chunk_str('abcdef', 0, ' ')
with six.assertRaisesRegex(self, ValueError, 'n=-42'):
chunk_str('abcdef', -42, ' ')
def test_uneven_split(self):
self.assertEqual(chunk_str('abcdef', 5, '|'), 'abcde|f')
self.assertEqual(chunk_str('abcdef', 4, '|'), 'abcd|ef')
self.assertEqual(chunk_str('abcdefg', 3, ' - '), 'abc - def - g')
def test_no_split(self):
self.assertEqual(chunk_str('', 2, '\n'), '')
self.assertEqual(chunk_str('a', 100, '\n'), 'a')
self.assertEqual(chunk_str('abcdef', 42, '|'), 'abcdef')
