def testPropertyNameIsLowercased(self):
"""
The property name must be lower-cased in the result.
"""
read = AARead('id', '')
self.assertTrue('hydropathy' in
clustersForSequence(read, ['HYDROPATHY']))
def testPropertyNameIsLowercased(self):
"""
The property name must be lower-cased in the result.
"""
read = AARead('id', '')
self.assertTrue(
'hydropathy' in clustersForSequence(read, ['HYDROPATHY']))
def testOnePropertyEmptySequence(self):
"""
If one property is wanted but the sequence is empty, a dict with the
property must be returned, and have an empty list value.
"""
read = AARead('id', '')
self.assertEqual({
'hydropathy': [],
}, clustersForSequence(read, ['hydropathy']))
def testOneProperty(self):
"""
If one property is wanted, a dict with the property must be returned,
and have the expected property cluster number.
"""
read = AARead('id', 'AI')
self.assertEqual({
'hydropathy': [3, 4],
}, clustersForSequence(read, ['hydropathy']))
def testTwoProperties(self):
"""
If two properties are wanted, a dict with the properties must be
returned, and have the expected property cluster numbers.
"""
read = AARead('id', 'AI')
self.assertEqual({
'composition': [1, 1],
'hydropathy': [3, 4],
}, clustersForSequence(read, ['composition', 'hydropathy']))
def testMissingAminoAcid(self):
"""
If an unknown amino acid appears in the sequence, its property cluster
must be the default (0).
"""
read = AARead('id', 'XX')
self.assertEqual({
'composition': [0, 0],
'hydropathy': [0, 0],
}, clustersForSequence(read, ['composition', 'hydropathy']))
def testOneProperty(self):
"""
If one property is wanted, a dict with the property must be returned,
and have the expected property cluster number.
"""
read = AARead('id', 'AI')
self.assertEqual(
{
'hydropathy': [3, 4],
},
clustersForSequence(read, ['hydropathy']))
def testOnePropertyEmptySequence(self):
"""
If one property is wanted but the sequence is empty, a dict with the
property must be returned, and have an empty list value.
"""
read = AARead('id', '')
self.assertEqual(
{
'hydropathy': [],
},
clustersForSequence(read, ['hydropathy']))
def plotAAClusters(sequence, propertyNames, showLines=True, showFigure=True):
"""
Plot amino acid property cluster numbers for a sequence.
@param sequence: An C{AARead} (or a subclass) instance.
@param propertyNames: An iterable of C{str} property names (each of which
must be a key of a key in the C{dark.aa.PROPERTY_CLUSTERS} C{dict}).
@param showLines: If C{True}, lines will be drawn between successive AA
property values. If not, just the values will be plotted as a scatter
plot (this greatly reduces visual clutter if the sequence is long and
AA property values are variable).
@param showFigure: If C{True}, display the plot. Passing C{False} is useful
in testing.
@raise ValueError: If an unknown property is given in C{propertyNames}.
@return: The return value from calling dark.aa.clustersForSequence:
a C{dict} keyed by (lowercase) property name, with values that are
C{list}s of the corresponding property value according to sequence
position.
"""
MISSING_AA_VALUE = 0
propertyClusters = clustersForSequence(sequence,
propertyNames,
missingAAValue=MISSING_AA_VALUE)
if showFigure:
minCluster = 1
maxCluster = -1
legend = []
x = np.arange(0, len(sequence))
plot = plt.plot if showLines else plt.scatter
for index, propertyName in enumerate(propertyClusters):
color = TABLEAU20[index]
clusterNumbers = propertyClusters[propertyName]
plot(x, clusterNumbers, color=color)
legend.append(patches.Patch(color=color, label=propertyName))
propertyMinCluster = min(clusterNumbers)
if propertyMinCluster < minCluster:
minCluster = propertyMinCluster
propertyMaxCluster = max(clusterNumbers)
if propertyMaxCluster > maxCluster:
maxCluster = propertyMaxCluster
plt.legend(handles=legend, loc=(0, 1.1))
plt.xlim(-0.2, len(sequence) - 0.8)
plt.ylim(minCluster - 0.5, maxCluster + 0.5)
plt.yticks(range(maxCluster + 1))
plt.xlabel('Sequence index')
plt.ylabel('Property cluster number')
plt.title(sequence.id)
plt.show()
return propertyClusters
def plotAAClusters(sequence, propertyNames, showLines=True, showFigure=True):
"""
Plot amino acid property cluster numbers for a sequence.
@param sequence: An C{AARead} (or a subclass) instance.
@param propertyNames: An iterable of C{str} property names (each of which
must be a key of a key in the C{dark.aa.PROPERTY_CLUSTERS} C{dict}).
@param showLines: If C{True}, lines will be drawn between successive AA
property values. If not, just the values will be plotted as a scatter
plot (this greatly reduces visual clutter if the sequence is long and
AA property values are variable).
@param showFigure: If C{True}, display the plot. Passing C{False} is useful
in testing.
@raise ValueError: If an unknown property is given in C{propertyNames}.
@return: The return value from calling dark.aa.clustersForSequence:
a C{dict} keyed by (lowercase) property name, with values that are
C{list}s of the corresponding property value according to sequence
position.
"""
MISSING_AA_VALUE = 0
propertyClusters = clustersForSequence(sequence, propertyNames,
missingAAValue=MISSING_AA_VALUE)
if showFigure:
minCluster = 1
maxCluster = -1
legend = []
x = np.arange(0, len(sequence))
plot = plt.plot if showLines else plt.scatter
for index, propertyName in enumerate(propertyClusters):
color = TABLEAU20[index]
clusterNumbers = propertyClusters[propertyName]
plot(x, clusterNumbers, color=color)
legend.append(patches.Patch(color=color, label=propertyName))
propertyMinCluster = min(clusterNumbers)
if propertyMinCluster < minCluster:
minCluster = propertyMinCluster
propertyMaxCluster = max(clusterNumbers)
if propertyMaxCluster > maxCluster:
maxCluster = propertyMaxCluster
plt.legend(handles=legend, loc=(0, 1.1))
plt.xlim(-0.2, len(sequence) - 0.8)
plt.ylim(minCluster - 0.5, maxCluster + 0.5)
plt.yticks(range(maxCluster + 1))
plt.xlabel('Sequence index')
plt.ylabel('Property cluster number')
plt.title(sequence.id)
plt.show()
return propertyClusters
def testMissingAminoAcid(self):
"""
If an unknown amino acid appears in the sequence, its property cluster
must be the default (0).
"""
read = AARead('id', 'XX')
self.assertEqual(
{
'composition': [0, 0],
'hydropathy': [0, 0],
},
clustersForSequence(read, ['composition', 'hydropathy']))
def testTwoProperties(self):
"""
If two properties are wanted, a dict with the properties must be
returned, and have the expected property cluster numbers.
"""
read = AARead('id', 'AI')
self.assertEqual(
{
'composition': [1, 1],
'hydropathy': [3, 4],
},
clustersForSequence(read, ['composition', 'hydropathy']))
def testMissingAminoAcidWithNonDefaultMissingValue(self):
"""
If an unknown amino acid appears in the sequence, its property cluster
must be the missing AA value that was passed.
"""
read = AARead('id', 'XX')
self.assertEqual({
'composition': [10, 10],
'hydropathy': [10, 10],
},
clustersForSequence(read,
['composition', 'hydropathy'],
missingAAValue=10))
def testMissingAminoAcidWithNonDefaultMissingValue(self):
"""
If an unknown amino acid appears in the sequence, its property cluster
must be the missing AA value that was passed.
"""
read = AARead('id', 'XX')
self.assertEqual(
{
'composition': [10, 10],
'hydropathy': [10, 10],
},
clustersForSequence(read, ['composition', 'hydropathy'],
missingAAValue=10))
def testDuplicatedPropertyName(self):
"""
If a property name is mentioned more than once, a dict with the
wanted properties must be returned, and have the expected property
cluster numbers.
"""
read = AARead('id', 'AI')
self.assertEqual(
{
'composition': [1, 1],
'hydropathy': [3, 4],
},
clustersForSequence(read, ['composition',
'hydropathy', 'hydropathy']))
def testDuplicatedPropertyName(self):
"""
If a property name is mentioned more than once, a dict with the
wanted properties must be returned, and have the expected property
cluster numbers.
"""
read = AARead('id', 'AI')
self.assertEqual(
{
'composition': [1, 1],
'hydropathy': [3, 4],
},
clustersForSequence(read,
['composition', 'hydropathy', 'hydropathy']))
def testNoProperties(self):
"""
If no properties are wanted, an empty dict must be returned.
"""
read = AARead('id', 'RRR')
self.assertEqual({}, clustersForSequence(read, []))
def testNoProperties(self):
"""
If no properties are wanted, an empty dict must be returned.
"""
read = AARead('id', 'RRR')
self.assertEqual({}, clustersForSequence(read, []))
