def testSmallDecode(self):
"""
Test modified fraction estimation in detection mode around a known modification in lambda
"""
# First methlyated A in lambda:
# strand motif onTarget seqid tpl
# 0 GCACNNNNNNGTT On 1 14983
start = 14900
end = 15100
referenceWindow = ReferenceWindow(1, "lambda_NEB3011", start, end)
bounds = (start, end)
self.kw._prepForReferenceWindow(referenceWindow)
kinetics = self.kw._summarizeReferenceRegion(bounds, True, False)
# Verify that we detect m6A mods at 14982 and 14991
m6AMods = [{
'frac': x['frac'],
'fracLow': x['fracLow'],
'fracUp': x['fracUp'],
'tpl': x['tpl'],
'strand': x['strand']
} for x in kinetics if 'frac' in x and x['tpl'] in (14982, 14991)]
print(m6AMods)
for mod in m6AMods:
self.assertGreater(mod["frac"], 0.5)
def test_private_api(self):
start = 50
end = 100
REF_GROUP_ID = "gi|12057207|gb|AE001439.1|"
referenceWindow = ReferenceWindow(0, REF_GROUP_ID, start, end)
bounds = (start, end)
rir = list(
self.kw.caseCmpH5.readsInRange(referenceWindow.refName,
referenceWindow.start,
referenceWindow.end))
self.assertEqual(len(rir), 301)
chunks = self.kw._fetchChunks(REF_GROUP_ID, (start, end),
self.kw.caseCmpH5)
factor = 1.0 / self.ds.readGroupTable[0].FrameRate
rawIpds = self.kw._loadRawIpds(rir, start, end, factor)
logging.critical(len(rawIpds))
# XXX note that this is very dependent on the exact order of reads
# found by readsInRange(), which may be altered by changes to the
# implementation of the dataset API. It should, however, remain
# consistent across equivalent input types.
# XXX 2015-08-28 disabling this for now because it will change if the
# dataset contains multiple .bam files
#self.assertEqual("%.4f" % rawIpds[0][2], "0.2665")
log.info(rawIpds)
chunks = self.kw._chunkRawIpds(rawIpds)
def test_small_decode(self):
"""Test for known modifications near the start of H. pylori genome"""
# XXX should have mods on 60- (m4C), 89+ (m6A), 91- (m6A)
start = 50
end = 100
REF_GROUP_ID = "gi|12057207|gb|AE001439.1|"
referenceWindow = ReferenceWindow(0, REF_GROUP_ID, start, end)
bounds = (start, end)
self.kw._prepForReferenceWindow(referenceWindow)
kinetics = self.kw._summarizeReferenceRegion(bounds, False, True)
mods = self.kw._decodePositiveControl(kinetics, bounds)
log.info(mods)
# Verify that we detect m6A mods at 14982 and 14991
m6AMods = [
x for x in mods
if x['modification'] == 'm6A' and x['tpl'] in (88, 90)
]
self.assertEqual(len(m6AMods), 2)
m4CMods = [
x for x in mods
if x['modification'] == 'm4C' and x['tpl'] in (59, )
]
self.assertEqual(len(m4CMods), 1)
for x in mods:
if x['strand'] == 0:
self.assertEqual(x['tpl'], 88)
else:
self.assertTrue(x['tpl'] in [59, 90])
def test_small_decode(self):
start = 985
end = 1065
REF_GROUP_ID = "gi|12057207|gb|AE001439.1|"
referenceWindow = ReferenceWindow(0, REF_GROUP_ID, start, end)
bounds = (start, end)
self.kw._prepForReferenceWindow(referenceWindow)
kinetics = self.kw._summarizeReferenceRegion(bounds, False, True)
mods = self.kw._decodePositiveControl(kinetics, bounds)
self.assertEqual(len(mods), 4)
def test_small_decode(self):
start = 50
end = 100
REF_GROUP_ID = "gi|12057207|gb|AE001439.1|"
referenceWindow = ReferenceWindow(0, REF_GROUP_ID, start, end)
bounds = (start, end)
self.kw._prepForReferenceWindow(referenceWindow)
kinetics = self.kw._summarizeReferenceRegion(bounds, False, True)
# XXX note that this is very dependent on the exact order of reads
# found by readsInRange(), which may be altered by changes to the
# implementation of the dataset API. It should be immune to stochastic
# effects, however.
self.assertEqual("%.5f" % kinetics[0]['ipdRatio'], "1.06460")
mods = self.kw._decodePositiveControl(kinetics, bounds)
self.assertEqual(len(mods), 3)
def testSmallDecode(self):
"""
Test a modification decode around a known modification in lambda
"""
# First methlyated A in lambda:
# strand motif onTarget seqid tpl
# 0 GCACNNNNNNGTT On 1 14983
start = 14900
end = 15100
referenceWindow = ReferenceWindow(1, "lambda_NEB3011", start, end)
bounds = (start, end)
self.kw._prepForReferenceWindow(referenceWindow)
kinetics = self.kw._summarizeReferenceRegion(bounds, False, True)
mods = self.kw._decodePositiveControl(kinetics, bounds)
print(mods)
# Verify that we detect m6A mods at 14982 and 14991
m6AMods = [x for x in mods if x['modification'] == 'm6A' and x['tpl'] in (14982, 14991)]
self.assertEqual(len(m6AMods), 2)