def test_noop__no_intervals(self):
seq = Seq(''.join([random.choice('ATGC') for i in range(200)]),
generic_dna)
orig_seq_record = SeqRecord(seq)
seq_record = copy.deepcopy(orig_seq_record)
interval_list = []
automated_intergenic_gc_fixer(seq_record, interval_list)
self.assertEqual(str(orig_seq_record.seq), str(seq_record.seq))
def test_noop__small_interval(self):
# Sequence is all GC so no changes to make.
seq = Seq(''.join([random.choice('GC') for i in range(200)]),
generic_dna)
orig_seq_record = SeqRecord(seq)
seq_record = copy.deepcopy(orig_seq_record)
interval_list = [(100, 125)]
automated_intergenic_gc_fixer(seq_record, interval_list)
self.assertEqual(str(orig_seq_record.seq), str(seq_record.seq))
def test_change_some(self):
"""Keep changing positions until target is met.
"""
# Repeat to make sure not stochastic result
# ...though it still could be :)
for i in range(100):
# Sequence is all AT so some will have to change.
seq = Seq(''.join([random.choice('AT') for i in range(200)]),
generic_dna)
orig_seq_record = SeqRecord(seq)
seq_record = copy.deepcopy(orig_seq_record)
self.assertTrue(GC(seq_record.seq) == 0)
interval_list = [(100, 125)]
automated_intergenic_gc_fixer(seq_record, interval_list)
for center in range(100, 125):
window_seq = seq_record.seq[center - 50:center + 50]
self.assertTrue(
GC(window_seq) >= 30,
"GC for window centered at %d is %f" %
(center, GC(window_seq)))
self.assertNotEqual(str(orig_seq_record.seq), str(seq_record.seq))
def test_change_all(self):
"""Change all positions around intervals.
"""
# Sequence is all AT so some will have to change.
seq = Seq(''.join([random.choice('AT') for i in range(200)]),
generic_dna)
orig_seq_record = SeqRecord(seq)
seq_record = copy.deepcopy(orig_seq_record)
self.assertTrue(GC(seq_record.seq) < 30)
interval_list = [(100, 125)]
constraint_obj = GCContentConstraints()
constraint_obj.local_window_lower_bound = 1.0
constraint_obj.local_window_upper_bound = 1.1 # no upper bound
automated_intergenic_gc_fixer(seq_record,
interval_list,
gc_content_constraint_obj=constraint_obj)
for center in range(100, 125):
window_seq = seq_record.seq[center - 50:center + 50]
self.assertTrue(
GC(window_seq) == 100, "GC for window centered at %d is %f" %
(center, GC(window_seq)))
self.assertNotEqual(str(orig_seq_record.seq), str(seq_record.seq))
def test_avoid_changes_in_shadows(self):
"""Avoid changing bases in CDS features.
"""
# Sequence is all AT so some will have to change.
seq = Seq(''.join([random.choice('AT') for i in range(200)]),
generic_dna)
orig_seq_record = SeqRecord(seq)
# We will expect a 60 base shadow from 60 - 120. This is the CDS
# and 20 bases upstream.
feature_1_loc = FeatureLocation(80, 120, strand=1)
feature_1 = SeqFeature(feature_1_loc, type='CDS', id=1)
orig_seq_record.features.append(feature_1)
seq_record = copy.deepcopy(orig_seq_record)
self.assertTrue(GC(seq_record.seq) == 0)
# Hit just one position.
interval_list = [(100, 101)]
# Aim for 100% to hit all bases possible.
constraint_obj = GCContentConstraints()
constraint_obj.local_window_lower_bound = 1.0
constraint_obj.local_window_upper_bound = 1.1 # no upper bound
automated_intergenic_gc_fixer(seq_record,
interval_list,
gc_content_constraint_obj=constraint_obj)
# Expect window centered at 100 to have GC 60, which is all
window_seq = seq_record.seq[50:150]
self.assertEqual(40, GC(window_seq))
# Make sure shadow seq is unchanged.
self.assertEqual(str(feature_1.extract(orig_seq_record.seq)),
str(feature_1.extract(seq_record.seq)))