def test_count(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with open(self.fasta(utils.SEQUENCES)) as fasta_handle:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.count([fasta_handle], profile_handle, 8)
utils.test_profile_file(filename, counts, 8)
def test_smooth(self):
# See test_kdistlib.test_ProfileDistance_dynamic_smooth
counts_left = Counter([
'AC', 'AG', 'AT', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT',
'TA', 'TG', 'TT'
])
counts_right = Counter([
'AC', 'AT', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT', 'TA',
'TC', 'TG', 'TT'
])
filename_left = self.empty()
filename_right = self.empty()
with utils.open_profile(self.profile(counts_left, 2)) as handle_left:
with utils.open_profile(self.profile(counts_right,
2)) as handle_right:
with utils.open_profile(filename_left, 'w') as out_left:
with utils.open_profile(filename_right, 'w') as out_right:
kmer.smooth(handle_left,
handle_right,
out_left,
out_right,
summary='min')
counts_left = Counter([
'AA', 'AA', 'AA', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT',
'TA', 'TA', 'TA'
])
counts_right = Counter([
'AA', 'AA', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT', 'TA',
'TA', 'TA', 'TA'
])
utils.test_profile_file(filename_left, counts_left, 2)
utils.test_profile_file(filename_right, counts_right, 2)
def test_scale(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename_left = self.empty()
filename_right = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename_left, 'w') as out_left:
with utils.open_profile(filename_right, 'w') as out_right:
kmer.scale(handle_left, handle_right, out_left, out_right)
if sum(counts_left.values()) < sum(counts_right.values()):
scale_left = sum(counts_right.values()) / sum(counts_left.values())
scale_right = 1.0
else:
scale_left = 1.0
scale_right = sum(counts_left.values()) / sum(counts_right.values())
for s in counts_left:
counts_left[s] *= scale_left
for s in counts_right:
counts_right[s] *= scale_right
utils.test_profile_file(filename_left, counts_left, 8)
utils.test_profile_file(filename_right, counts_right, 8)
def test_scale(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename_left = self.empty()
filename_right = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename_left, 'w') as out_left:
with utils.open_profile(filename_right, 'w') as out_right:
kmer.scale(handle_left, handle_right, out_left, out_right)
if sum(counts_left.values()) < sum(counts_right.values()):
scale_left = sum(counts_right.values()) / sum(counts_left.values())
scale_right = 1.0
else:
scale_left = 1.0
scale_right = sum(counts_left.values()) / sum(counts_right.values())
for s in counts_left:
counts_left[s] *= scale_left
for s in counts_right:
counts_right[s] *= scale_right
utils.test_profile_file(filename_left, counts_left, 8)
utils.test_profile_file(filename_right, counts_right, 8)
def test_count(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with open(self.fasta(utils.SEQUENCES)) as fasta_handle:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.count([fasta_handle], profile_handle, 8)
utils.test_profile_file(filename, counts, 8)
def test_convert(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with open(self.profile_old_format(counts, 8)) as handle:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.convert([handle], profile_handle)
utils.test_profile_file(filename, counts, 8)
def test_convert(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with open(self.profile_old_format(counts, 8)) as handle:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.convert([handle], profile_handle)
utils.test_profile_file(filename, counts, 8)
def test_balance(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts, 8)) as input_handle:
with utils.open_profile(filename, 'w') as output_handle:
kmer.balance(input_handle, output_handle)
counts.update(dict((utils.reverse_complement(s), c) for s, c in counts.items()))
utils.test_profile_file(filename, counts, 8)
def test_balance(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts, 8)) as input_handle:
with utils.open_profile(filename, 'w') as output_handle:
kmer.balance(input_handle, output_handle)
counts.update(dict((utils.reverse_complement(s), c) for s, c in counts.items()))
utils.test_profile_file(filename, counts, 8)
def test_count_by_record(self):
counts_by_record = [utils.counts(record, 8) for record in utils.SEQUENCES]
names = [str(i) for i, _ in enumerate(counts_by_record)]
filename = self.empty()
with open(self.fasta(utils.SEQUENCES, names=names)) as fasta_handle:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.count([fasta_handle], profile_handle, 8, by_record=True)
for name, counts in zip(names, counts_by_record):
utils.test_profile_file(filename, counts, 8, name=name)
def test_profile_save(self):
counts = utils.counts(utils.SEQUENCES, 4)
profile = klib.Profile(utils.as_array(counts, 4))
filename = self.empty()
with utils.open_profile(filename, 'w') as profile_handle:
profile.save(profile_handle)
utils.test_profile_file(filename, counts, 4)
def test_profile_save(self):
counts = utils.counts(utils.SEQUENCES, 4)
profile = klib.Profile(utils.as_array(counts, 4))
filename = self.empty()
with utils.open_profile(filename, 'w') as profile_handle:
profile.save(profile_handle)
utils.test_profile_file(filename, counts, 4)
def test_merge(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.merge(handle_left, handle_right, profile_handle)
utils.test_profile_file(filename, counts_left + counts_right, 8)
def test_merge(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.merge(handle_left, handle_right, profile_handle)
utils.test_profile_file(filename, counts_left + counts_right, 8)
def test_count_multi(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with open(self.fasta(utils.SEQUENCES_LEFT)) as handle_left:
with open(self.fasta(utils.SEQUENCES_RIGHT)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.count([handle_left, handle_right], profile_handle, 8, names=['a', 'b'])
utils.test_profile_file(filename, counts_left, 8, name='a')
utils.test_profile_file(filename, counts_right, 8, name='b')
def test_cat_prefixes(self):
counts_a = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_b = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_a, 8, name='X')) as handle_a:
with utils.open_profile(self.profile(counts_b, 8, name='X')) as handle_b:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.cat([handle_a, handle_b], profile_handle, prefixes=['a_', 'b_'])
utils.test_profile_file(filename, counts_a, 8, name='a_X')
utils.test_profile_file(filename, counts_b, 8, name='b_X')
def test_cat_prefixes(self):
counts_a = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_b = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_a, 8, name='X')) as handle_a:
with utils.open_profile(self.profile(counts_b, 8, name='X')) as handle_b:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.cat([handle_a, handle_b], profile_handle, prefixes=['a_', 'b_'])
utils.test_profile_file(filename, counts_a, 8, name='a_X')
utils.test_profile_file(filename, counts_b, 8, name='b_X')
def test_shrink(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts, 8)) as input_handle:
with utils.open_profile(filename, 'w') as output_handle:
kmer.shrink(input_handle, output_handle, 1)
counts = Counter(dict((t, sum(counts[u] for u in counts
if u.startswith(t)))
for t in set(s[:-1] for s in counts)))
utils.test_profile_file(filename, counts, 7)
def test_shrink(self):
counts = utils.counts(utils.SEQUENCES, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts, 8)) as input_handle:
with utils.open_profile(filename, 'w') as output_handle:
kmer.shrink(input_handle, output_handle, 1)
counts = Counter(dict((t, sum(counts[u] for u in counts
if u.startswith(t)))
for t in set(s[:-1] for s in counts)))
utils.test_profile_file(filename, counts, 7)
def test_shuffle(self):
# See test_klib.profile_shuffle
counts = utils.counts(utils.SEQUENCES, 2)
filename = self.empty()
with utils.open_profile(self.profile(counts, 2)) as input_handle:
with utils.open_profile(filename, 'w') as output_handle:
np.random.seed(100)
kmer.shuffle(input_handle, output_handle)
counts = dict(zip([''.join(s) for s in itertools.product('ACGT', repeat=2)],
[13, 7, 6, 18, 12, 1, 13, 17, 16, 12, 23, 27, 24, 17, 18, 12]))
utils.test_profile_file(filename, counts, 2)
def test_shuffle(self):
# See test_klib.profile_shuffle
counts = utils.counts(utils.SEQUENCES, 2)
filename = self.empty()
with utils.open_profile(self.profile(counts, 2)) as input_handle:
with utils.open_profile(filename, 'w') as output_handle:
np.random.seed(100)
kmer.shuffle(input_handle, output_handle)
counts = dict(zip([''.join(s) for s in itertools.product('ACGT', repeat=2)],
[13, 7, 6, 18, 12, 1, 13, 17, 16, 12, 23, 27, 24, 17, 18, 12]))
utils.test_profile_file(filename, counts, 2)
def test_count_multi_by_record(self):
counts_by_record_left = [utils.counts(record, 8) for record in utils.SEQUENCES_LEFT]
counts_by_record_right = [utils.counts(record, 8) for record in utils.SEQUENCES_RIGHT]
names_left = [str(i) for i, _ in enumerate(counts_by_record_left)]
names_right = [str(i) for i, _ in enumerate(counts_by_record_right)]
filename = self.empty()
with open(self.fasta(utils.SEQUENCES_LEFT, names=names_left)) as handle_left:
with open(self.fasta(utils.SEQUENCES_RIGHT, names=names_right)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.count([handle_left, handle_right], profile_handle, 8, names=['a', 'b'], by_record=True)
for name, counts in zip(names_left, counts_by_record_left):
utils.test_profile_file(filename, counts, 8, name='a_' + name)
for name, counts in zip(names_right, counts_by_record_right):
utils.test_profile_file(filename, counts, 8, name='b_' + name)
def test_merge_custom_name(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.merge(handle_left, handle_right, profile_handle, custom_merger='numpy.multiply')
counts_mult = Counter(dict((s, counts_left[s] * counts_right[s])
for s in set(counts_left) & set(counts_right)))
utils.test_profile_file(filename, counts_mult, 8)
def test_merge_custom_name(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.merge(handle_left, handle_right, profile_handle, custom_merger='numpy.multiply')
counts_mult = Counter(dict((s, counts_left[s] * counts_right[s])
for s in set(counts_left) & set(counts_right)))
utils.test_profile_file(filename, counts_mult, 8)
def test_merge_custom_expr(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.merge(handle_left, handle_right, profile_handle, custom_merger='(left + right) * np.logical_xor(left, right)')
counts_xor = counts_left + counts_right
for s in set(counts_left) & set(counts_right):
del counts_xor[s]
utils.test_profile_file(filename, counts_xor, 8)
def test_merge_custom_expr(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename, 'w') as profile_handle:
kmer.merge(handle_left, handle_right, profile_handle, custom_merger='(left + right) * np.logical_xor(left, right)')
counts_xor = counts_left + counts_right
for s in set(counts_left) & set(counts_right):
del counts_xor[s]
utils.test_profile_file(filename, counts_xor, 8)
def test_positive(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename_left = self.empty()
filename_right = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename_left, 'w') as out_left:
with utils.open_profile(filename_right, 'w') as out_right:
kmer.positive(handle_left, handle_right, out_left, out_right)
utils.test_profile_file(filename_left, Counter(s for s in counts_left.elements()
if s in counts_right), 8)
utils.test_profile_file(filename_right, Counter(s for s in counts_right.elements()
if s in counts_left), 8)
def test_positive(self):
counts_left = utils.counts(utils.SEQUENCES_LEFT, 8)
counts_right = utils.counts(utils.SEQUENCES_RIGHT, 8)
filename_left = self.empty()
filename_right = self.empty()
with utils.open_profile(self.profile(counts_left, 8)) as handle_left:
with utils.open_profile(self.profile(counts_right, 8)) as handle_right:
with utils.open_profile(filename_left, 'w') as out_left:
with utils.open_profile(filename_right, 'w') as out_right:
kmer.positive(handle_left, handle_right, out_left, out_right)
utils.test_profile_file(filename_left, Counter(s for s in counts_left.elements()
if s in counts_right), 8)
utils.test_profile_file(filename_right, Counter(s for s in counts_right.elements()
if s in counts_left), 8)
def test_smooth(self):
# See test_kdistlib.test_ProfileDistance_dynamic_smooth
counts_left = Counter(['AC', 'AG', 'AT', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT', 'TA', 'TG', 'TT'])
counts_right = Counter(['AC', 'AT', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT', 'TA', 'TC', 'TG', 'TT'])
filename_left = self.empty()
filename_right = self.empty()
with utils.open_profile(self.profile(counts_left, 2)) as handle_left:
with utils.open_profile(self.profile(counts_right, 2)) as handle_right:
with utils.open_profile(filename_left, 'w') as out_left:
with utils.open_profile(filename_right, 'w') as out_right:
kmer.smooth(handle_left, handle_right, out_left, out_right, summary='min')
counts_left = Counter(['AA', 'AA', 'AA', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT', 'TA', 'TA', 'TA'])
counts_right = Counter(['AA', 'AA', 'CA', 'CC', 'CG', 'CT', 'GA', 'GC', 'GG', 'GT', 'TA', 'TA', 'TA', 'TA'])
utils.test_profile_file(filename_left, counts_left, 2)
utils.test_profile_file(filename_right, counts_right, 2)
