class TestCallPatterns(unittest.TestCase):
"""The following motivates these tests:
BetaRat(5.5, 0.5, 11.0, 16.0) cdf: 0.00392333077990058 map: 3.27321043801743 ppf: 2.47393902966
BetaRat(4.5, 1.5, 6.0, 21.0) cdf: 0.00570220397191901 map: 3.475475215347 ppf: 1.98195901549
As you can see, different methods can make different calls."""
def setUp(self):
ref_seq = Seq.Seq('GA' * 9 + 'GC' * 6 + 'GT' * 15)
query_seq = 'AA' * 4 + 'GA' * 5 + 'AC' + 'GC' * 5 + 'GT' * 15
aln_string = '>seq1\n{0}'.format(query_seq)
self.aln = Alignment(helpers.parse_fasta(aln_string).values(),
reference_sequence=ref_seq)
self.mutation_patterns = [mut_pattern.GA, mut_pattern.GM]
def test_map_caller(self):
for result in self.aln.multiple_context_analysis(
self.mutation_patterns, caller='map'):
call = result['call']
self.assertEqual(call['call_pattern'], 'GA')
def test_ppf_caller(self):
for result in self.aln.multiple_context_analysis(
self.mutation_patterns,
caller='q_0.05',
quants=[0.05],
pos_quants_only=False):
call = result['call']
self.assertEqual(call['call_pattern'], 'GM')
def test_cdf_caller(self):
"""Note that CDF should be evaluated as smaller => more extreme, whereas the other statistics are the
inverse"""
for result in self.aln.multiple_context_analysis(
self.mutation_patterns, caller='cutoff_cdf'):
call = result['call']
self.assertEqual(call['call_pattern'], 'GM')
def test_ppf_caller_without_all_quants(self):
"""Should raise if we don't have the given statistic to test for call comparison."""
analysis = self.aln.multiple_context_analysis(self.mutation_patterns,
caller='q_0.05')
with self.assertRaises(ValueError):
analysis.next()
def test_not_calling_negatives(self):
"""We should only call sequences when they are actually positive, even if there are non-positive
sequences with more extreme call statistics."""
for result in self.aln.multiple_context_analysis(
self.mutation_patterns, caller='map',
significance_level=0.005):
call = result['call']
self.assertEqual(call['call_pattern'], 'GM')
class TestCallPatterns(unittest.TestCase):
"""The following motivates these tests:
BetaRat(5.5, 0.5, 11.0, 16.0) cdf: 0.00392333077990058 map: 3.27321043801743 ppf: 2.47393902966
BetaRat(4.5, 1.5, 6.0, 21.0) cdf: 0.00570220397191901 map: 3.475475215347 ppf: 1.98195901549
As you can see, different methods can make different calls."""
def setUp(self):
ref_seq = Seq.Seq('GA'*9 + 'GC'*6 + 'GT'*15)
query_seq = 'AA'*4 + 'GA'*5 + 'AC' + 'GC'*5 + 'GT'*15
aln_string = '>seq1\n{0}'.format(query_seq)
self.aln = Alignment(helpers.parse_fasta(aln_string).values(), reference_sequence=ref_seq)
self.mutation_patterns=[mut_pattern.GA, mut_pattern.GM]
def test_map_caller(self):
for result in self.aln.multiple_context_analysis(self.mutation_patterns, caller='map'):
call = result['call']
self.assertEqual(call['call_pattern'], 'GA')
def test_ppf_caller(self):
for result in self.aln.multiple_context_analysis(self.mutation_patterns, caller='q_0.05',
quants=[0.05], pos_quants_only=False):
call = result['call']
self.assertEqual(call['call_pattern'], 'GM')
def test_cdf_caller(self):
"""Note that CDF should be evaluated as smaller => more extreme, whereas the other statistics are the
inverse"""
for result in self.aln.multiple_context_analysis(self.mutation_patterns, caller='cutoff_cdf'):
call = result['call']
self.assertEqual(call['call_pattern'], 'GM')
def test_ppf_caller_without_all_quants(self):
"""Should raise if we don't have the given statistic to test for call comparison."""
analysis = self.aln.multiple_context_analysis(self.mutation_patterns, caller='q_0.05')
with self.assertRaises(ValueError):
analysis.next()
def test_not_calling_negatives(self):
"""We should only call sequences when they are actually positive, even if there are non-positive
sequences with more extreme call statistics."""
for result in self.aln.multiple_context_analysis(self.mutation_patterns, caller='map',
significance_level=0.005):
call = result['call']
self.assertEqual(call['call_pattern'], 'GM')
