def test_init_varied_input(self):
# init as string
b = BiologicalSequence('ACCGGXZY')
self.assertEqual(str(b), 'ACCGGXZY')
self.assertEqual(b.id, "")
self.assertEqual(b.description, "")
# init as string with optional values
b = BiologicalSequence('ACCGGXZY', 'test-seq-1',
'The first test sequence')
self.assertEqual(str(b), 'ACCGGXZY')
self.assertEqual(b.id, "test-seq-1")
self.assertEqual(b.description, "The first test sequence")
# test init as a different string
b = BiologicalSequence('WRRTY')
self.assertEqual(str(b), 'WRRTY')
# init as list
b = BiologicalSequence(list('ACCGGXZY'))
self.assertEqual(str(b), 'ACCGGXZY')
self.assertEqual(b.id, "")
self.assertEqual(b.description, "")
# init as tuple
b = BiologicalSequence(tuple('ACCGGXZY'))
self.assertEqual(str(b), 'ACCGGXZY')
self.assertEqual(b.id, "")
self.assertEqual(b.description, "")
def setUp(self):
"""Setup for Fasta tests."""
self.strings = ['AAAA', 'CCCC', 'gggg', 'uuuu']
self.fasta_no_label = '>0\nAAAA\n>1\nCCCC\n>2\ngggg\n>3\nuuuu'
self.fasta_with_label =\
'>1st\nAAAA\n>2nd\nCCCC\n>3rd\nGGGG\n>4th\nUUUU'
self.fasta_with_label_lw2 =\
'>1st\nAA\nAA\n>2nd\nCC\nCC\n>3rd\nGG\nGG\n>4th\nUU\nUU'
self.alignment_dict = {
'1st': 'AAAA',
'2nd': 'CCCC',
'3rd': 'GGGG',
'4th': 'UUUU'
}
self.sequence_objects_a = [
DNASequence('ACTCGAGATC', 'seq1'),
DNASequence('GGCCT', 'seq2')
]
self.sequence_objects_b = [
BiologicalSequence('ACTCGAGATC', 'seq1'),
BiologicalSequence('GGCCT', 'seq2')
]
seqs = [
DNASequence("ACC--G-GGTA..", id="seq1"),
DNASequence("TCC--G-GGCA..", id="seqs2")
]
self.alignment = Alignment(seqs)
def test_make_nr_foundation_alignment_few(self):
result = _make_nr_foundation_alignment(self.foundation_alignment,
self.extension_genus_dic_few)
self.assertEqual(list(result), [
BiologicalSequence("AAA---", id="PBB1", description="Phoma"),
BiologicalSequence("AAG---", id="CBB1", description="Candida"),
])
def test_fasta_with_many_fungi(self):
result = fungi_from_fasta(self.fasta_many_fungi, self.accession,
self.taxonomy_with_fungi)
self.assertEqual(list(result), [
BiologicalSequence("GGGG", id="AB123", description="Fungi"),
BiologicalSequence("CCCC", id="AB125", description="Fungi"),
BiologicalSequence("AAAA", id="AB126", description="Fungi"),
])
def test_eq(self):
self.assertTrue(self.b1 == self.b1)
self.assertTrue(self.b2 == self.b2)
self.assertTrue(self.b3 == self.b3)
self.assertTrue(self.b1 != self.b3)
self.assertTrue(self.b1 != self.b2)
self.assertTrue(self.b2 != self.b3)
# identicial sequences of the same type are equal, even if they have
# different ids and/or descriptions
self.assertTrue(
BiologicalSequence('ACGT') == BiologicalSequence('ACGT'))
self.assertTrue(
BiologicalSequence('ACGT', id='a') == BiologicalSequence('ACGT',
id='b'))
self.assertTrue(
BiologicalSequence('ACGT', description='c') == BiologicalSequence(
'ACGT', description='d'))
self.assertTrue(
BiologicalSequence('ACGT', id='a', description='c') ==
BiologicalSequence('ACGT', id='b', description='d'))
# different type causes sequences to not be equal
self.assertFalse(
BiologicalSequence('ACGT') == NucleotideSequence('ACGT'))
def generator():
yield BiologicalSequence('ACGT',
id='',
description='',
quality=range(4))
yield RNASequence('GAU', id=' foo \t\t bar ', description='')
yield DNASequence('TAG', id='', description='foo\n\n bar\n')
yield BiologicalSequence('A',
id='foo',
description='bar baz',
quality=[42])
def test_traceback(self):
score_m = [[0, -5, -7, -9], [-5, 2, -3, -5], [-7, -3, 4, -1],
[-9, -5, -1, 6], [-11, -7, -3, 1]]
score_m = np.array(score_m)
tback_m = [[0, 3, 3, 3], [2, 1, 3, 3], [2, 2, 1, 3], [2, 2, 2, 1],
[2, 2, 2, 2]]
tback_m = np.array(tback_m)
# start at bottom-right
expected = ([BiologicalSequence("ACG-")], [BiologicalSequence("ACGT")],
1, 0, 0)
actual = _traceback(tback_m, score_m, Alignment([DNA('ACG')]),
Alignment([DNA('ACGT')]), 4, 3)
self.assertEqual(actual, expected)
# four sequences in two alignments
score_m = [[0, -5, -7, -9], [-5, 2, -3, -5], [-7, -3, 4, -1],
[-9, -5, -1, 6], [-11, -7, -3, 1]]
score_m = np.array(score_m)
tback_m = [[0, 3, 3, 3], [2, 1, 3, 3], [2, 2, 1, 3], [2, 2, 2, 1],
[2, 2, 2, 2]]
tback_m = np.array(tback_m)
# start at bottom-right
expected = ([BiologicalSequence("ACG-"),
BiologicalSequence("ACG-")],
[BiologicalSequence("ACGT"),
BiologicalSequence("ACGT")], 1, 0, 0)
actual = _traceback(tback_m, score_m,
Alignment([DNA('ACG', 's1'),
DNA('ACG', 's2')]),
Alignment([DNA('ACGT', 's3'),
DNA('ACGT', 's4')]), 4, 3)
self.assertEqual(actual, expected)
# start at highest-score
expected = ([BiologicalSequence("ACG")], [BiologicalSequence("ACG")],
6, 0, 0)
actual = _traceback(tback_m, score_m, Alignment([DNA('ACG')]),
Alignment([DNA('ACGT')]), 3, 3)
self.assertEqual(actual, expected)
# terminate traceback before top-right
tback_m = [[0, 3, 3, 3], [2, 1, 3, 3], [2, 2, 0, 3], [2, 2, 2, 1],
[2, 2, 2, 2]]
tback_m = np.array(tback_m)
expected = ("G", "G", 6, 2, 2)
expected = ([BiologicalSequence("G")], [BiologicalSequence("G")], 6, 2,
2)
actual = _traceback(tback_m, score_m, Alignment([DNA('ACG')]),
Alignment([DNA('ACGT')]), 3, 3)
self.assertEqual(actual, expected)
def test_k_word_frequencies(self):
# overlapping = True
expected = defaultdict(int)
expected['A'] = 3 / 7.
expected['C'] = 1 / 7.
expected['G'] = 1 / 7.
expected['T'] = 2 / 7.
self.assertEqual(self.b1.k_word_frequencies(1, overlapping=True),
expected)
expected = defaultdict(int)
expected['GAT'] = 1 / 5.
expected['ATT'] = 1 / 5.
expected['TTA'] = 1 / 5.
expected['TAC'] = 1 / 5.
expected['ACA'] = 1 / 5.
self.assertEqual(self.b1.k_word_frequencies(3, overlapping=True),
expected)
# overlapping = False
expected = defaultdict(int)
expected['GAT'] = 1 / 2.
expected['TAC'] = 1 / 2.
self.assertEqual(self.b1.k_word_frequencies(3, overlapping=False),
expected)
expected = defaultdict(int)
expected['GATTACA'] = 1.0
self.assertEqual(self.b1.k_word_frequencies(7, overlapping=False),
expected)
expected = defaultdict(int)
empty = BiologicalSequence('')
self.assertEqual(empty.k_word_frequencies(1, overlapping=False),
expected)
def test_k_words(self):
# overlapping = True
self.assertEqual(list(self.b1.k_words(1, overlapping=True)),
['G', 'A', 'T', 'T', 'A', 'C', 'A'])
self.assertEqual(list(self.b1.k_words(2, overlapping=True)),
['GA', 'AT', 'TT', 'TA', 'AC', 'CA'])
self.assertEqual(list(self.b1.k_words(3, overlapping=True)),
['GAT', 'ATT', 'TTA', 'TAC', 'ACA'])
self.assertEqual(list(self.b1.k_words(7, overlapping=True)),
['GATTACA'])
self.assertEqual(list(self.b1.k_words(8, overlapping=True)), [])
# overlapping = False
self.assertEqual(list(self.b1.k_words(1, overlapping=True)),
['G', 'A', 'T', 'T', 'A', 'C', 'A'])
self.assertEqual(list(self.b1.k_words(2, overlapping=False)),
['GA', 'TT', 'AC'])
self.assertEqual(list(self.b1.k_words(3, overlapping=False)),
['GAT', 'TAC'])
self.assertEqual(list(self.b1.k_words(7, overlapping=False)),
['GATTACA'])
self.assertEqual(list(self.b1.k_words(8, overlapping=False)), [])
# error on invalid k
self.assertRaises(ValueError, list, self.b1.k_words(0))
self.assertRaises(ValueError, list, self.b1.k_words(-42))
# tests with different sequences
self.assertEqual(list(self.b8.k_words(3, overlapping=False)),
['HE.', '.--', '..L'])
b = BiologicalSequence('')
self.assertEqual(list(b.k_words(3)), [])
def test_init_with_validation(self):
self.assertRaises(BiologicalSequenceError,
BiologicalSequence,
"ACC",
validate=True)
try:
# no error raised when only allow characters are passed
BiologicalSequence("..--..", validate=True)
except BiologicalSequenceError:
self.assertTrue(False)
def test_distance(self):
# note that test_hamming_distance covers default behavior more
# extensively
self.assertEqual(self.b1.distance(self.b1), 0.0)
self.assertEqual(self.b1.distance(BiologicalSequence('GATTACC')),
1. / 7)
def dumb_distance(x, y):
return 42
self.assertEqual(self.b1.distance(self.b1, distance_fn=dumb_distance),
42)
def setUp(self):
self.b1 = BiologicalSequence('GATTACA')
self.b2 = BiologicalSequence('ACCGGTACC',
id="test-seq-2",
description="A test sequence")
self.b3 = BiologicalSequence('GREG',
id="test-seq-3",
description="A protein sequence")
self.b4 = BiologicalSequence('PRTEIN', id="test-seq-4")
self.b5 = BiologicalSequence('LLPRTEIN',
description="some description")
self.b6 = BiologicalSequence('ACGTACGTACGT')
self.b7 = BiologicalSequence('..--..')
self.b8 = BiologicalSequence('HE..--..LLO')
def test_gap_maps(self):
# in sequence with no gaps, the gap_maps are identical
self.assertEqual(self.b1.gap_maps(),
([0, 1, 2, 3, 4, 5, 6], [0, 1, 2, 3, 4, 5, 6]))
# in sequence with all gaps, the map of degapped to gapped is the empty
# list (bc its length is 0), and the map of gapped to degapped is all
# None
self.assertEqual(self.b7.gap_maps(),
([], [None, None, None, None, None, None]))
self.assertEqual(self.b8.gap_maps(),
([0, 1, 8, 9, 10],
[0, 1, None, None, None, None, None, None, 2, 3, 4]))
# example from the gap_maps doc string
self.assertEqual(
BiologicalSequence('-ACCGA-TA-').gap_maps(),
([1, 2, 3, 4, 5, 7, 8], [None, 0, 1, 2, 3, 4, None, 5, 6, None]))
def _coerce_alignment_input_type(seq, disallow_alignment):
""" Converts variety of types into an skbio.Alignment object
"""
if isinstance(seq, string_types):
return Alignment([BiologicalSequence(seq)])
elif isinstance(seq, BiologicalSequence):
return Alignment([seq])
elif isinstance(seq, Alignment):
if disallow_alignment:
# This will disallow aligning either a pair of alignments, or an
# alignment and a sequence. We don't currently support this for
# local alignment as there is not a clear usecase, and it's also
# not exactly clear how this would work.
raise TypeError("Aligning alignments is not currently supported "
"with the aligner function that you're calling.")
else:
return seq
else:
raise TypeError("Unsupported type provided to aligner: %r." %
type(seq))
def demux_sequences(sequences, barcodes, barcode_map):
yield BiologicalSequence('')
def gen():
for c in components:
yield BiologicalSequence(c[2],
id=c[0],
description=c[1],
quality=c[3])
def gen():
yield BiologicalSequence('ACGT',
id='foo',
description='bar',
quality=range(4))
yield BiologicalSequence('ACG', id='foo', description='bar')
def test_fasta_with_fungi(self):
result = fungi_from_fasta(self.fasta_with_fungi, self.accession,
self.taxonomy_with_fungi)
self.assertEqual(
list(result),
[BiologicalSequence("ATCG", id="AB21", description="Fungi")])
def test_getitem(self):
self.assertEqual(self.b1[0], BiologicalSequence('G'))
self.assertEqual(self.b1[:], BiologicalSequence('GATTACA'))
self.assertEqual(self.b1[::-1], BiologicalSequence('ACATTAG'))
def missing_qual_gen():
for seq in (RNASequence('A',
quality=[42]), BiologicalSequence('AG'),
DNASequence('GG', quality=[41, 40])):
yield seq
def test_degap(self):
self.assertEqual(self.b1.degap(), self.b1)
self.assertEqual(self.b7.degap(), BiologicalSequence(''))
self.assertEqual(self.b8.degap(), BiologicalSequence('HELLO'))
def blank_seq_gen():
for seq in (DNASequence('A'), BiologicalSequence(''),
RNASequence('GG')):
yield seq
def test_fraction_diff(self):
self.assertEqual(self.b1.fraction_diff(self.b1), 0., 5)
self.assertEqual(self.b1.fraction_diff(BiologicalSequence('GATTACC')),
1. / 7., 5)
def test_fraction_same(self):
self.assertAlmostEqual(self.b1.fraction_same(self.b1), 1., 5)
self.assertAlmostEqual(
self.b1.fraction_same(BiologicalSequence('GATTACC')), 6. / 7., 5)
def _traceback(traceback_matrix,
score_matrix,
aln1,
aln2,
start_row,
start_col,
gap_character='-'):
# cache some values for simpler
aend = _traceback_encoding['alignment-end']
match = _traceback_encoding['match']
vgap = _traceback_encoding['vertical-gap']
hgap = _traceback_encoding['horizontal-gap']
# initialize the result alignments
aln1_sequence_count = aln1.sequence_count()
aligned_seqs1 = [[] for e in range(aln1_sequence_count)]
aln2_sequence_count = aln2.sequence_count()
aligned_seqs2 = [[] for e in range(aln2_sequence_count)]
current_row = start_row
current_col = start_col
best_score = score_matrix[current_row, current_col]
current_value = None
while current_value != aend:
current_value = traceback_matrix[current_row, current_col]
if current_value == match:
for aligned_seq, input_seq in zip(aligned_seqs1, aln1):
aligned_seq.append(str(input_seq[current_col - 1]))
for aligned_seq, input_seq in zip(aligned_seqs2, aln2):
aligned_seq.append(str(input_seq[current_row - 1]))
current_row -= 1
current_col -= 1
elif current_value == vgap:
for aligned_seq in aligned_seqs1:
aligned_seq.append('-')
for aligned_seq, input_seq in zip(aligned_seqs2, aln2):
aligned_seq.append(str(input_seq[current_row - 1]))
current_row -= 1
elif current_value == hgap:
for aligned_seq, input_seq in zip(aligned_seqs1, aln1):
aligned_seq.append(str(input_seq[current_col - 1]))
for aligned_seq in aligned_seqs2:
aligned_seq.append('-')
current_col -= 1
elif current_value == aend:
continue
else:
raise ValueError("Invalid value in traceback matrix: %s" %
current_value)
for i in range(aln1_sequence_count):
aligned_seq = ''.join(aligned_seqs1[i][::-1])
seq_id = _get_seq_id(aln1[i], str(i))
aligned_seqs1[i] = BiologicalSequence(aligned_seq, id=seq_id)
for i in range(aln2_sequence_count):
aligned_seq = ''.join(aligned_seqs2[i][::-1])
seq_id = _get_seq_id(aln2[i], str(i + aln1_sequence_count))
aligned_seqs2[i] = BiologicalSequence(aligned_seq, id=seq_id)
return (aligned_seqs1, aligned_seqs2, best_score, current_col, current_row)
def quality_filter_sequences(sequences, barcodes):
return BiologicalSequence('')
