def test_WildtypeFilter(self):
f = data_query.WildtypeFilter()
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='')))
self.assertEqual(f.score(None, ov), 1)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='Delta gene-01')))
self.assertEqual(f.score(None, ov), -1)
def test_filter_observed_results(self):
gen = self.ConcreteDataQueryGenerator()
gen.filters = [
data_query.TemperatureRangeFilter(min=36., max=38.),
]
observed_results = [
data_model.ObservedValue(metadata=data_model.ObservedResultMetadata(environment=data_model.Environment(temperature=36.5))),
data_model.ObservedValue(metadata=data_model.ObservedResultMetadata(environment=data_model.Environment(temperature=35.0))),
data_model.ObservedValue(metadata=data_model.ObservedResultMetadata(environment=data_model.Environment(temperature=37.0))),
]
result = gen.filter_observed_results(None, observed_results)
self.assertEqual(set(result.observed_results), set([observed_results[0], observed_results[2]]))
def test_score(self):
ov = [
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=6))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=6))),
]
f = [
data_query.TemperatureRangeFilter(min=36.5, max=37.5),
data_query.PhRangeFilter(min=6.5, max=7.5),
data_query.TemperatureNormalFilter(mean=37., std=1),
]
runner = data_query.FilterRunner(f[0])
self.assertEqual(list(runner.score(None, ov).ravel()), [1., 1., -1., -1., 1., 1., -1., -1.])
runner = data_query.FilterRunner(f[1])
self.assertEqual(list(runner.score(None, ov).ravel()), [1., -1., 1., -1., 1., -1., 1., -1.])
runner = data_query.FilterRunner(f[2])
s = 2 * scipy.stats.norm.cdf(-1)
numpy.testing.assert_almost_equal(list(runner.score(None, ov).ravel()), [1., 1., s, s, 1., 1., s, s], decimal=5)
def test_TaxonomicDistanceFilter(self):
# example 1
f = data_query.TaxonomicDistanceFilter('Mycoplasma pneumoniae M129')
self.assertEqual(f.max, 8.)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Mycoplasma pneumoniae M129')))
self.assertEqual(f.score(None, ov), 1.)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Mycoplasma pneumoniae')))
self.assertEqual(f.scale, 8./5)
self.assertEqual(f.score(None, ov), math.exp(-1/f.scale))
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Escherichia coli')))
self.assertEqual(f.score(None, ov), math.exp(-5))
# example 2
f = data_query.TaxonomicDistanceFilter('Mycoplasma pneumoniae M129', max=5)
self.assertEqual(f.max, 5.)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Mycoplasma pneumoniae M129')))
self.assertEqual(f.score(None, ov), 1.)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Mycoplasma pneumoniae')))
self.assertEqual(f.score(None, ov), math.exp(-1/f.scale))
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Escherichia coli')))
self.assertEqual(f.score(None, ov), -1)
# example 3
f = data_query.TaxonomicDistanceFilter('Mycoplasma genitalium')
self.assertEqual(f.max, 7.)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Mycoplasma genitalium G37')))
self.assertEqual(f.score(None, ov), 1.)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Mycoplasma genitalium')))
self.assertEqual(f.score(None, ov), 1)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(taxon='Mycoplasma')))
self.assertEqual(f.score(None, ov), math.exp(-1/f.scale))
def test_run(self):
ov = [
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=6))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37.1, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37.1, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37.01, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37.01, ph=6))),
]
f = [
data_query.TemperatureNormalFilter(mean=37, std=1),
data_query.PhRangeFilter(min=6.5, max=7.5),
]
runner = data_query.FilterRunner(f)
# return_info=True
result = runner.run(None, ov, return_info=True)
self.assertEqual(result.all_observed_results, ov)
s1 = 2 * scipy.stats.norm.cdf(-1)
s01 = 2 * scipy.stats.norm.cdf(-0.1)
s001 = 2 * scipy.stats.norm.cdf(-0.01)
numpy.testing.assert_almost_equal(result.all_scores, numpy.array([[1., 1., s1, s1, s01, s01, s001, s001],
[1., -1., 1., -1., 1., -1., 1., -1.]], ndmin=2).transpose())
self.assertEqual(result.observed_results, [ov[0], ov[6], ov[4], ov[2]])
numpy.testing.assert_almost_equal(result.scores, numpy.array([[1., s001, s01, s1], [1., 1., 1., 1.]], ndmin=2).transpose())
self.assertEqual(result.observed_value_indices, [0, 6, 4, 2])
# return_info=False
ordered_observed_results = runner.run(None, ov, return_info=False)
self.assertEqual(ordered_observed_results, [ov[0], ov[6], ov[4], ov[2]])
def test_order(self):
ov = [
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=6))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=6))),
]
# one filter
s = numpy.array([0.9, -1, 0.5, 1, -1.1, 0, -1.3, 0.3], ndmin=2).transpose()
runner = data_query.FilterRunner([])
order_o, order_s, i_order_o = runner.order(ov, s)
self.assertEqual(order_o, [ov[3], ov[0], ov[2], ov[7], ov[5], ov[1], ov[4], ov[6]])
numpy.testing.assert_equal(order_s, numpy.array([1, 0.9, 0.5, 0.3, 0, -1, -1.1, -1.3], ndmin=2).transpose())
self.assertEqual(i_order_o, [3, 0, 2, 7, 5, 1, 4, 6])
# multiple filters
s = numpy.array([[0.9, -1, 0.5, 1, -1.1, 0, -1.3, 0.3], [0, 0, 0, 0, 0, 0, 0, 0]], ndmin=2).transpose()
runner = data_query.FilterRunner([])
order_o, order_s, i_order_o = runner.order(ov, s)
self.assertEqual(order_o, [ov[3], ov[0], ov[2], ov[7], ov[5], ov[1], ov[4], ov[6]])
numpy.testing.assert_equal(order_s, numpy.array(
[[1, 0.9, 0.5, 0.3, 0, -1, -1.1, -1.3], [0, 0, 0, 0, 0, 0, 0, 0]], ndmin=2).transpose())
self.assertEqual(i_order_o, [3, 0, 2, 7, 5, 1, 4, 6])
def test_filter(self):
ov = [
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=6))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=7))),
data_model.ObservedValue(value=1, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=37, ph=6))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=7))),
data_model.ObservedValue(value=2, metadata=data_model.ObservedResultMetadata(
environment=data_model.Environment(temperature=36, ph=6))),
]
# one filter
s = numpy.array([1, -1, 0.5, 1, -1, 0, -1, 0.3], ndmin=2).transpose()
runner = data_query.FilterRunner([])
filt_o, filt_s, i_filt_o = runner.filter(ov, s)
self.assertEqual(filt_o, [ov[0], ov[2], ov[3], ov[5], ov[7]])
numpy.testing.assert_equal(filt_s, numpy.array([1, 0.5, 1, 0, 0.3], ndmin=2).transpose())
self.assertEqual(i_filt_o, [0, 2, 3, 5, 7])
# multiple filters
s = numpy.array([[1, -1, 0.5, 1, -1, 0, -1, 0.3], [0, 0, 0, 0, 0, 0, 0, -1]], ndmin=2).transpose()
runner = data_query.FilterRunner([])
filt_o, filt_s, i_filt_o = runner.filter(ov, s)
self.assertEqual(filt_o, [ov[0], ov[2], ov[3], ov[5]])
numpy.testing.assert_equal(filt_s, numpy.array([[1, 0.5, 1, 0], [0, 0, 0, 0]], ndmin=2).transpose())
self.assertEqual(i_filt_o, [0, 2, 3, 5])
def test_OptionsFilter(self):
f = data_query.OptionsFilter(('metadata', 'genetics', 'variation', ), [''])
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='')))
self.assertEqual(f.score(None, ov), 1)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='wildtype')))
self.assertEqual(f.score(None, ov), -1)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='Delta gene-01')))
self.assertEqual(f.score(None, ov), -1)
f = data_query.OptionsFilter(('metadata', 'genetics', 'variation', ), ['wildtype'])
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='')))
self.assertEqual(f.score(None, ov), -1)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='wildtype')))
self.assertEqual(f.score(None, ov), 1)
ov = data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(genetics=data_model.Genetics(variation='Delta gene-01')))
self.assertEqual(f.score(None, ov), -1)
def metadata_dump(self, component):
""" Calculate a consensus statistical representation of the one or more observed values
Args:
component (:obj:`models.Observation`): model component dump data for
Returns:
:obj:`list` of :obj:`data_model.ObservedResultMetadata`: data model metadata object
"""
genetics = None
environment = None
cross_references = []
method = None
synonym = []
meta = component._metadata
taxon = meta.taxon[0].name if meta.taxon else None
variation = meta.cell_line[0].name if meta.cell_line else None
genetics = data_model.Genetics(taxon=taxon, variation=variation)
temperature = meta.conditions[
0].temperature if meta.conditions else None
ph = meta.conditions[0].ph if meta.conditions else None
media = meta.conditions[0].media if meta.conditions else None
growth_status = meta.conditions[
0].growth_status if meta.conditions else None
growth_system = meta.conditions[
0].growth_system if meta.conditions else None
environment = data_model.Environment(temperature=temperature,
ph=ph,
media=media,
growth_status=growth_status,
growth_system=growth_system)
name = meta.method[0].name if meta.method else None
description = meta.method[0].comments if meta.method else None
performer = meta.method[0].performer if meta.method else None
hardware = meta.method[0].hardware if meta.method else None
software = meta.method[0].software if meta.method else None
method = data_model.Method(name=name,
description=description,
performer=performer,
hardware=hardware,
software=software)
if meta.resource:
for item in meta.resource:
cross_references.append(
data_model.Resource(namespace=item.namespace, id=item._id))
else:
cross_references.append(
data_model.Resource(namespace=None, id=None))
if meta.synonym:
for item in meta.synonym:
synonym.append(data_model.Synonym(name=item.name))
else:
synonym.append(data_model.Synonym(name=None))
metadata_result = data_model.ObservedResultMetadata(
genetics=genetics,
environment=environment,
cross_references=cross_references,
method=method,
synonym=synonym)
return metadata_result
def obs_val(temperature):
return data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(environment=data_model.Environment(temperature=temperature)))
def obs_val(ph):
return data_model.ObservedValue(
metadata=data_model.ObservedResultMetadata(environment=data_model.Environment(ph=ph)))
def test_ObservedResultMetadata(self):
o = data_model.ObservedResultMetadata()
o.genetics = data_model.Genetics(taxon='Mycoplasma pneumoniae',
variation='ΔMPN001')
o.environment = data_model.Environment(temperature=37,
ph=7.,
media='Hayflick')
o.reference = data_model.Reference(title='title',
author='author',
year=2017,
volume=1,
number=1,
pages='1-10')
o.method = data_model.ExperimentalMethod(
name='assay', description='description of assay')
observable = data_model.Observable(
interaction=data_model.Reaction(id='AtpSynthase'),
specie=data_model.Specie(id='ATP'),
compartment=data_model.Compartment(id='c'),
property='K_m',
)
ov = data_model.ObservedValue(
observable=observable,
value=1.0,
error=0.5,
units='U/mg',
)
o.observed_result.append(ov)
o.validate()
self.assertEqual(o.observed_result, [ov])
self.assertEqual(ov.metadata, o)
self.assertEqual(ov.observable.interaction.id, 'AtpSynthase')
self.assertEqual(ov.observable.specie.id, 'ATP')
self.assertEqual(ov.observable.compartment.id, 'c')
self.assertEqual(ov.observable.property, 'K_m')
self.assertEqual(ov.value, 1.0)
self.assertEqual(ov.error, 0.5)
self.assertEqual(ov.units, 'U/mg')
self.assertEqual(o.genetics.taxon, 'Mycoplasma pneumoniae')
self.assertEqual(o.genetics.variation, 'ΔMPN001')
self.assertEqual(o.environment.temperature, 37.)
self.assertEqual(o.environment.ph, 7.)
self.assertEqual(o.environment.media, 'Hayflick')
self.assertEqual(o.reference.title, 'title')
self.assertEqual(o.reference.author, 'author')
self.assertEqual(o.reference.year, 2017)
self.assertEqual(o.reference.volume, 1)
self.assertEqual(o.reference.number, 1)
self.assertEqual(o.reference.pages, '1-10')
self.assertEqual(o.method.name, 'assay')
self.assertEqual(o.method.description, 'description of assay')
