def test_filter_results(self):
"""
Tests filter results
good for regression tests, need to do better verification...
"""
peak1 = Peak("chr15", 1, 10, "ENSG1", .04 , "-", 50, 60, 1, 52, .04, 32, 0)
peak2 = Peak("chr15", 200, 300, "ENSG2", .06 , "-", 140, 160, 2, 239, .06, 45, 0)
results = [{'loc': ['chr15', 'ENSG00000198901', 91509274, 91537804, '-'],
'Nclusters': 24,
'nreads': 2086,
'threshold': 32,
'clusters': [peak1, peak2]}]
transcriptome_size = 10000
transcriptome_reads = 100000
result = filter_results(results, .07, transcriptome_size, transcriptome_reads, False, False, "foo")
self.assertSetEqual(set(['chr15\t1\t10\tENSG1_1_52\t0.04\t-\t50\t60', 'chr15\t200\t300\tENSG2_2_239\t0.06\t-\t140\t160']), result)
result = filter_results(results, .05, transcriptome_size, transcriptome_reads, False, False)
self.assertSetEqual(set(['chr15\t1\t10\tENSG1_1_52\t0.04\t-\t50\t60']), result)
def test_broken_filter_results(self):
"""
Tests filter results, expects no output, even from transcriptome wide estimations
"""
peak1 = Peak("chr15", 1, 10, "ENSG1", .04, "-", 50, 60, 1, 52, .04, 32,
0)
peak2 = Peak("chr15", 200, 300, "ENSG2", .06, "-", 140, 160, 2, 239,
.06, 45, 0)
results = [{
'loc': ['chr15', 'ENSG00000198901', 91509274, 91537804, '-'],
'Nclusters': 24,
'nreads': 2086,
'threshold': 32,
'clusters': [peak1, peak2]
}]
transcriptome_size = 10000
transcriptome_reads = 100000
#I had my mental model of how the global cutoff was should have worked wrong the entire time...
result = filter_results(results, .07, transcriptome_size,
transcriptome_reads, True, False, "foo")
self.assertSetEqual(
set([
'chr15\t1\t10\tENSG1_1_52\t0.04\t-\t50\t60',
'chr15\t200\t300\tENSG2_2_239\t0.06\t-\t140\t160'
]), result)
def test_bonferroni_correct_filter_results(self):
"""
Tests to make sure bonferroni correction works
"""
transcriptome_size = 10000
transcriptome_reads = 100000
peak1 = Peak("chr15", 1, 10, "ENSG1", .04, "-", 50, 60, 1, 52, .04, 32,
0)
peak2 = Peak("chr15", 200, 300, "ENSG2", .06, "-", 140, 160, 2, 239,
.06, 45, 0)
results = [{
'loc': ['chr15', 'ENSG00000198901', 91509274, 91537804, '-'],
'Nclusters': 24,
'nreads': 2086,
'threshold': 32,
'clusters': [peak1, peak2]
}]
result = filter_results(results, .09, transcriptome_size,
transcriptome_reads, False, True)
self.assertSetEqual(set(['chr15\t1\t10\tENSG1_1_52\t0.08\t-\t50\t60']),
result)
def test_transcriptome_filter(self):
"""
Tests transcriptome filter
not great tests, but good enough to make sure we don't have regressions
"""
cluster = Peak(0, 0, 0, 0, 0, 0, 0, 0, 0, 5, 0, 10, 0, 0, 0, 0)
#cluster = {'Nreads' : 5, "size" : 10}
transcriptome_size = 1000
transcriptome_reads = 10000
poisson_cutoff = .05
result = transcriptome_filter(poisson_cutoff, transcriptome_size,
transcriptome_reads, cluster)
self.assertEqual(result, 1)
#cluster = {'Nreads' : 10000, "size" : 100}
cluster = Peak(0, 0, 0, 0, 0, 0, 0, 0, 0, 10000, 0, 100, 0, 0, 0, 0)
transcriptome_size = 1000
transcriptome_reads = 10000
poisson_cutoff = .05
result = transcriptome_filter(poisson_cutoff, transcriptome_size,
transcriptome_reads, cluster)
self.assertEqual(result, 0.0)
#cluster = {'Nreads' : 0, "size" : 0}
cluster = Peak(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
transcriptome_size = 0
transcriptome_reads = 10000
poisson_cutoff = .05
result = transcriptome_filter(poisson_cutoff, transcriptome_size,
transcriptome_reads, cluster)
self.assertEqual(result, 1)