def testGetTrackSubsetTS(self):
subsetField1Value2 = FlatTracksTS()
subsetField1Value2['B'] = self.t2
subsetField1Value2['C'] = self.t3
self._assertEqualTrackStructure(
subsetField1Value2,
self.flatTrackStructure.getTrackSubsetTS('field 1', 'value 2'))
subsetField2val6 = FlatTracksTS()
subsetField2val6['B'] = self.t2
self._assertEqualTrackStructure(
subsetField2val6,
self.flatTrackStructure.getTrackSubsetTS('field 2', '6'))
subsetField3None = FlatTracksTS()
subsetField3None['C'] = self.t3
self._assertEqualTrackStructure(
subsetField3None,
self.flatTrackStructure.getTrackSubsetTS('field 3', 'None'))
empty = FlatTracksTS()
self._assertEqualTrackStructure(
empty,
self.flatTrackStructure.getTrackSubsetTS('field does not exist',
'value'))
self._assertEqualTrackStructure(
empty,
self.flatTrackStructure.getTrackSubsetTS('field 1',
'val does not exist'))
def testGetSplittedByCategoryTS(self):
splitByField1 = TrackStructureV2()
field1value1 = FlatTracksTS()
field1value2 = FlatTracksTS()
splitByField1['value 1'] = field1value1
splitByField1['value 1']['A'] = self.t1
splitByField1['value 2'] = field1value2
splitByField1['value 2']['B'] = self.t2
splitByField1['value 2']['C'] = self.t3
self._assertEqualTrackStructure(
splitByField1,
self.flatTrackStructure.getSplittedByCategoryTS('field 1'))
splitByField2 = TrackStructureV2()
field2val6 = FlatTracksTS()
splitByField2['6'] = field2val6
splitByField2['6']['B'] = self.t2
self._assertEqualTrackStructure(
splitByField2,
self.flatTrackStructure.getSplittedByCategoryTS('field 2'))
splitByField3 = TrackStructureV2()
field3None = FlatTracksTS()
splitByField3['None'] = field3None
splitByField3['None']['C'] = self.t3
self._assertEqualTrackStructure(
splitByField3,
self.flatTrackStructure.getSplittedByCategoryTS("field 3"))
empty = TrackStructureV2()
self._assertEqualTrackStructure(
empty,
self.flatTrackStructure.getSplittedByCategoryTS(
'field does not exist'))
def _buildTestTrees(self):
# inputTree splitOnA splitOnB pairwise (A vs B)
# / \ | / \ / \
# A B C D E t1_t2 t1_t3
# | / \ / \ /\ /\ / \ / \
# C D E A B A B A B Q R Q R
# | | | | / \ | | | | | | | |
# t1 t2 t3 t1 D E C t2 C t3 t1 t2 t1 t2
# | | | |
# t2 t3 t1 t1
self.t1 = SingleTrackTS(Track(['t1']), {'field 1': 'value 1'})
self.t2 = SingleTrackTS(Track(['t2']), {
'field 1': 'value 2',
'field 2': '6'
})
self.t3 = SingleTrackTS(Track(['t3']), {
'field 1': 'value 2',
'field 3': 'None'
})
self.inputTree = TrackStructureV2()
self.inputTree['A'] = TrackStructureV2()
self.inputTree['A']['C'] = self.t1
self.inputTree['B'] = TrackStructureV2()
self.inputTree['B']['D'] = self.t2
self.inputTree['B']['E'] = self.t3
# correct result of the input tree splitted on node A
self.splittedOnNodeA = TrackStructureV2()
self.splittedOnNodeA['C'] = TrackStructureV2()
self.splittedOnNodeA['C']['A'] = self.t1
self.splittedOnNodeA['C']['B'] = TrackStructureV2()
self.splittedOnNodeA['C']['B']['D'] = self.t2
self.splittedOnNodeA['C']['B']['E'] = self.t3
# correct result of the input tree splitted on node B
self.splittedOnNodeB = TrackStructureV2()
self.splittedOnNodeB['D'] = TrackStructureV2()
self.splittedOnNodeB['D']['A'] = TrackStructureV2()
self.splittedOnNodeB['D']['A']['C'] = self.t1
self.splittedOnNodeB['D']['B'] = self.t2
self.splittedOnNodeB['E'] = TrackStructureV2()
self.splittedOnNodeB['E']['A'] = TrackStructureV2()
self.splittedOnNodeB['E']['A']['C'] = self.t1
self.splittedOnNodeB['E']['B'] = self.t3
self.pairwiseCombinations = TrackStructureV2()
self.pairwiseCombinations["['t1']_['t2']"] = TrackStructureV2()
self.pairwiseCombinations["['t1']_['t2']"]['query'] = self.t1
self.pairwiseCombinations["['t1']_['t2']"]['reference'] = self.t2
self.pairwiseCombinations["['t1']_['t3']"] = TrackStructureV2()
self.pairwiseCombinations["['t1']_['t3']"]['query'] = self.t1
self.pairwiseCombinations["['t1']_['t3']"]['reference'] = self.t3
self.flatTrackStructure = FlatTracksTS()
self.flatTrackStructure['A'] = self.t1
self.flatTrackStructure['B'] = self.t2
self.flatTrackStructure['C'] = self.t3
def testGetFlattenedTS(self):
getFlattenedTsResult = FlatTracksTS()
# TODO Lonneke find better way for naming these
getFlattenedTsResult["['t1']"] = self.t1
getFlattenedTsResult["['t1'] (2)"] = self.t1
getFlattenedTsResult["['t2']"] = self.t2
getFlattenedTsResult["['t3']"] = self.t3
self._assertEqualTrackStructure(getFlattenedTsResult,
self.splittedOnNodeB.getFlattenedTS())
def getFlatTracksTS(genome, guiSelectedGSuite):
ts = FlatTracksTS()
gsuite = getGSuiteFromGalaxyTN(guiSelectedGSuite)
for gsTrack in gsuite.allTracks():
assert gsTrack.trackName is not None, "Gstrack name is None %s" % gsTrack
track = PlainTrack(gsTrack.trackName)
metadata = OrderedDict(title=gsTrack.title, genome=str(genome))
metadata.update(gsTrack.attributes)
assert track is not None
assert metadata is not None
ts[gsTrack.title] = SingleTrackTS(track, metadata)
return ts
def setUp(self):
stsRef1 = SingleTrackTS(Mock(spec=Track), dict(title="trackA"))
ts1 = FlatTracksTS()
ts1["reference"] = stsRef1
ts1.result = 0.5
ts1["query"] = Mock(spec=SingleTrackTS)
stsRef2 = SingleTrackTS(Mock(spec=Track), dict(title="trackB"))
ts2 = FlatTracksTS()
ts2["reference"] = stsRef2
ts2.result = 1.2
ts2["query"] = Mock(spec=SingleTrackTS)
self.resultTS = TrackStructureV2()
self.resultTS['0'] = ts1
self.resultTS['1'] = ts2
class TestTrackStructure(unittest.TestCase):
def _buildTestTrees(self):
# inputTree splitOnA splitOnB pairwise (A vs B)
# / \ | / \ / \
# A B C D E t1_t2 t1_t3
# | / \ / \ /\ /\ / \ / \
# C D E A B A B A B Q R Q R
# | | | | / \ | | | | | | | |
# t1 t2 t3 t1 D E C t2 C t3 t1 t2 t1 t2
# | | | |
# t2 t3 t1 t1
self.t1 = SingleTrackTS(Track(['t1']), {'field 1': 'value 1'})
self.t2 = SingleTrackTS(Track(['t2']), {
'field 1': 'value 2',
'field 2': '6'
})
self.t3 = SingleTrackTS(Track(['t3']), {
'field 1': 'value 2',
'field 3': 'None'
})
self.inputTree = TrackStructureV2()
self.inputTree['A'] = TrackStructureV2()
self.inputTree['A']['C'] = self.t1
self.inputTree['B'] = TrackStructureV2()
self.inputTree['B']['D'] = self.t2
self.inputTree['B']['E'] = self.t3
# correct result of the input tree splitted on node A
self.splittedOnNodeA = TrackStructureV2()
self.splittedOnNodeA['C'] = TrackStructureV2()
self.splittedOnNodeA['C']['A'] = self.t1
self.splittedOnNodeA['C']['B'] = TrackStructureV2()
self.splittedOnNodeA['C']['B']['D'] = self.t2
self.splittedOnNodeA['C']['B']['E'] = self.t3
# correct result of the input tree splitted on node B
self.splittedOnNodeB = TrackStructureV2()
self.splittedOnNodeB['D'] = TrackStructureV2()
self.splittedOnNodeB['D']['A'] = TrackStructureV2()
self.splittedOnNodeB['D']['A']['C'] = self.t1
self.splittedOnNodeB['D']['B'] = self.t2
self.splittedOnNodeB['E'] = TrackStructureV2()
self.splittedOnNodeB['E']['A'] = TrackStructureV2()
self.splittedOnNodeB['E']['A']['C'] = self.t1
self.splittedOnNodeB['E']['B'] = self.t3
self.pairwiseCombinations = TrackStructureV2()
self.pairwiseCombinations["['t1']_['t2']"] = TrackStructureV2()
self.pairwiseCombinations["['t1']_['t2']"]['query'] = self.t1
self.pairwiseCombinations["['t1']_['t2']"]['reference'] = self.t2
self.pairwiseCombinations["['t1']_['t3']"] = TrackStructureV2()
self.pairwiseCombinations["['t1']_['t3']"]['query'] = self.t1
self.pairwiseCombinations["['t1']_['t3']"]['reference'] = self.t3
self.flatTrackStructure = FlatTracksTS()
self.flatTrackStructure['A'] = self.t1
self.flatTrackStructure['B'] = self.t2
self.flatTrackStructure['C'] = self.t3
def setUp(self):
self._buildTestTrees()
def _assertEqualTrackStructure(self, correct, output):
self.assertEqual(correct.keys(), output.keys())
for key, value in correct.items():
self._assertEqualTrackStructure(correct[key], output[key])
self.assertIsInstance(output[key], correct[key].__class__)
self.assertIsInstance(output, TrackStructureV2)
if isinstance(correct[key], SingleTrackTS):
self.assertEqual(correct[key], output[key])
def testMakeTreeSegregatedByCategory(self):
# test splitting on a node that has a single category
singleCategoryOutput = self.inputTree.makeTreeSegregatedByCategory(
self.inputTree['A'])
self._assertEqualTrackStructure(singleCategoryOutput,
self.splittedOnNodeA)
# test splitting on a node that has multiple categories
singleCategoryOutput = self.inputTree.makeTreeSegregatedByCategory(
self.inputTree['B'])
self._assertEqualTrackStructure(singleCategoryOutput,
self.splittedOnNodeB)
# test splitting on a node without categories (should return an error)
with self.assertRaises(AssertionError):
self.inputTree.makeTreeSegregatedByCategory(
self.inputTree['A']['C'])
# TODO lonneke test with root as input
# should this return the same structure as the input?
# should metadata be moved around?
# should the new structure be different from input and have more levels?
def testMakePairwiseCombinations(self):
pairwiseOutput = self.inputTree['A'].makePairwiseCombinations(
self.inputTree['B'])
self.assertEqual(pairwiseOutput, self.pairwiseCombinations)
# combination between empty TrackStructures should result in just an empty TrackStructure
empty = TrackStructureV2()
empty.makePairwiseCombinations(empty)
self.assertEqual(empty.makePairwiseCombinations(empty), empty)
# TODO Lonneke add more border cases?
def testGetMetadataField(self):
self.assertItemsEqual(('field 1', 'field 2', 'field 3'),
self.flatTrackStructure.getMetadataFields())
def testGetAllValuesForMetadataField(self):
self.assertItemsEqual((
'value 1',
'value 2',
), self.flatTrackStructure.getAllValuesForMetadataField('field 1'))
self.assertItemsEqual(
('6', ),
self.flatTrackStructure.getAllValuesForMetadataField('field 2'))
self.assertItemsEqual(
('None', ),
self.flatTrackStructure.getAllValuesForMetadataField('field 3'))
def testGetFlattenedTS(self):
getFlattenedTsResult = FlatTracksTS()
# TODO Lonneke find better way for naming these
getFlattenedTsResult["['t1']"] = self.t1
getFlattenedTsResult["['t1'] (2)"] = self.t1
getFlattenedTsResult["['t2']"] = self.t2
getFlattenedTsResult["['t3']"] = self.t3
self._assertEqualTrackStructure(getFlattenedTsResult,
self.splittedOnNodeB.getFlattenedTS())
def testGetSplittedByCategoryTS(self):
splitByField1 = TrackStructureV2()
field1value1 = FlatTracksTS()
field1value2 = FlatTracksTS()
splitByField1['value 1'] = field1value1
splitByField1['value 1']['A'] = self.t1
splitByField1['value 2'] = field1value2
splitByField1['value 2']['B'] = self.t2
splitByField1['value 2']['C'] = self.t3
self._assertEqualTrackStructure(
splitByField1,
self.flatTrackStructure.getSplittedByCategoryTS('field 1'))
splitByField2 = TrackStructureV2()
field2val6 = FlatTracksTS()
splitByField2['6'] = field2val6
splitByField2['6']['B'] = self.t2
self._assertEqualTrackStructure(
splitByField2,
self.flatTrackStructure.getSplittedByCategoryTS('field 2'))
splitByField3 = TrackStructureV2()
field3None = FlatTracksTS()
splitByField3['None'] = field3None
splitByField3['None']['C'] = self.t3
self._assertEqualTrackStructure(
splitByField3,
self.flatTrackStructure.getSplittedByCategoryTS("field 3"))
empty = TrackStructureV2()
self._assertEqualTrackStructure(
empty,
self.flatTrackStructure.getSplittedByCategoryTS(
'field does not exist'))
def testGetTrackSubsetTS(self):
subsetField1Value2 = FlatTracksTS()
subsetField1Value2['B'] = self.t2
subsetField1Value2['C'] = self.t3
self._assertEqualTrackStructure(
subsetField1Value2,
self.flatTrackStructure.getTrackSubsetTS('field 1', 'value 2'))
subsetField2val6 = FlatTracksTS()
subsetField2val6['B'] = self.t2
self._assertEqualTrackStructure(
subsetField2val6,
self.flatTrackStructure.getTrackSubsetTS('field 2', '6'))
subsetField3None = FlatTracksTS()
subsetField3None['C'] = self.t3
self._assertEqualTrackStructure(
subsetField3None,
self.flatTrackStructure.getTrackSubsetTS('field 3', 'None'))
empty = FlatTracksTS()
self._assertEqualTrackStructure(
empty,
self.flatTrackStructure.getTrackSubsetTS('field does not exist',
'value'))
self._assertEqualTrackStructure(
empty,
self.flatTrackStructure.getTrackSubsetTS('field 1',
'val does not exist'))
def testIsPairedTs(self):
self.assertTrue(
self.pairwiseCombinations["['t1']_['t2']"].isPairedTs())
# the name of the tool.
# """
def create_track(file_name, trackName):
from gtrackcore.core.Api import importFile
importFile(file_name, genome="hg18", trackName=trackName)
t = PlainTrack([trackName])
single_track_ts = SingleTrackTS(t, {"title": trackName})
return single_track_ts
if __name__ == "__main__":
from gold.track.TrackStructure import SingleTrackTS, FlatTracksTS
from gold.track.Track import PlainTrack
track1 = create_track("testfile8.bed", "testfile8")
track2 = create_track("testfile7.bed", "testfile7")
#run_analysis("testfile.k")
#t = PlainTrack(["test"])
ts = FlatTracksTS()
ts["test1"] = track1
ts["test2"] = track2
print(ts)
analysisBins = UserBinSource("chr1", "*", genome="hg18")
RandomizationGuiTool.run_on_extracted_variables(ts, analysisBins, 1, WITHIN_TRACKS_CATEGORY,
PERMUTED_SEGS_AND_INTERSEGS_STR, galaxyFn="./testfile.gsuite", genome="hg18")
def execute(cls, choices, galaxyFn=None, username=''):
'''
Is called when execute-button is pushed by web-user. Should print
output as HTML to standard out, which will be directed to a results page
in Galaxy history. If getOutputFormat is anything else than HTML, the
output should be written to the file with path galaxyFn. If needed,
StaticFile can be used to get a path where additional files can be put
(e.g. generated image files). choices is a list of selections made by
web-user in each options box.
'''
import numpy
numpy.seterr(all='raise')
cls._setDebugModeIfSelected(choices)
genome = choices.genome
analysisQuestion = choices.analysisName
similaryStatClassName = choices.similarityFunc if choices.similarityFunc else GSuiteStatUtils.T5_RATIO_OF_OBSERVED_TO_EXPECTED_OVERLAP
summaryFunc = choices.summaryFunc if choices.summaryFunc else 'average'
reverse = 'Yes' if choices.reversed else 'No'
gsuite = getGSuiteFromGalaxyTN(choices.gsuite)
analysisBins = UserBinMixin.getUserBinSource(choices)
# tracks = [Track(x.trackName, trackTitle=x.title) for x in gsuite.allTracks()]
trackTitles = CommonConstants.TRACK_TITLES_SEPARATOR.join(
[quote(x.title, safe='') for x in gsuite.allTracks()])
import quick.gsuite.GuiBasedTsFactory as factory
ts = factory.getFlatTracksTS(genome=genome,
guiSelectedGSuite=choices.gsuite)
additionalResultsDict = OrderedDict()
additionalAttributesDict = OrderedDict()
if analysisQuestion in [cls.Q1, cls.Q2, cls.Q3]:
additionalAttributesDict = cls.getSelectedAttributesForEachTrackDict(
choices.additionalAttributes, gsuite)
# additional analysis
stats = [CountStat, CountElementStat]
additionalResultsDict = runMultipleSingleValStatsOnTracks(
ts, stats, analysisBins)
if analysisQuestion == cls.Q1:
analysisSpec = AnalysisSpec(MultitrackSummarizedInteractionV2Stat)
analysisSpec.addParameter('multitrackSummaryFunc', 'raw')
analysisSpec.addParameter(
'pairwiseStatistic', GSuiteStatUtils.
PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similaryStatClassName])
analysisSpec.addParameter(
'summaryFunc',
GSuiteStatUtils.SUMMARY_FUNCTIONS_MAPPER[summaryFunc])
analysisSpec.addParameter('reverse', reverse)
analysisSpec.addParameter('ascending', 'No')
analysisSpec.addParameter('trackTitles', trackTitles)
results = dictifyTSResult(
doAnalysis(analysisSpec, analysisBins,
ts).getGlobalResult()['Result'])
gsPerTrackResultsModel = GSuitePerTrackResultModel(
results,
['Similarity to rest of tracks in suite (%s)' % summaryFunc],
additionalResultsDict=additionalResultsDict,
additionalAttributesDict=additionalAttributesDict)
if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
columnTitles, decoratedResultsDict = \
gsPerTrackResultsModel.generateColumnTitlesAndResultsDict(choices.leadAttribute)
else:
columnTitles, decoratedResultsDict = \
gsPerTrackResultsModel.generateColumnTitlesAndResultsDict()
core = HtmlCore()
core.begin()
core.divBegin(divId='results-page')
core.divBegin(divClass='results-section')
core.header(analysisQuestion)
topTrackTitle = results.keys()[0]
core.paragraph('''
The track "%s" is the most representative track of the GSuite with %s %s similarity to the rest of the tracks
as measured by "%s" track similarity measure.
''' % (topTrackTitle, results[topTrackTitle], summaryFunc,
similaryStatClassName))
addTableWithTabularAndGsuiteImportButtons(
core,
choices,
galaxyFn,
cls.Q1_SHORT,
decoratedResultsDict,
columnTitles,
gsuite=gsuite,
results=results,
gsuiteAppendAttrs=['similarity_score'],
sortable=True)
# plot
columnInd = 0
if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
columnInd = 1
res = GSuiteTracksCoincidingWithQueryTrackTool.drawPlot(
results,
additionalResultsDict,
'Similarity to rest of tracks in suite (%s)' % summaryFunc,
columnInd=columnInd)
core.line(res)
core.divEnd()
core.divEnd()
core.end()
# elif analysisQuestion == cls.Q2:
# analysisSpec = AnalysisSpec(GSuiteRepresentativenessOfTracksRankingsWrapperStat)
# analysisSpec.addParameter('pairwiseStatistic', GSuiteStatUtils.PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similaryStatClassName])
# analysisSpec.addParameter('summaryFunc', GSuiteStatUtils.SUMMARY_FUNCTIONS_MAPPER[summaryFunc])
# analysisSpec.addParameter('reverse', reverse)
# analysisSpec.addParameter('ascending', 'Yes')
# analysisSpec.addParameter('trackTitles', trackTitles)
# results = doAnalysis(analysisSpec, analysisBins, tracks).getGlobalResult()
#
# gsPerTrackResultsModel = GSuitePerTrackResultModel(
# results, ['Similarity to rest of tracks in suite (%s)' % summaryFunc],
# additionalResultsDict=additionalResultsDict,
# additionalAttributesDict=additionalAttributesDict)
# if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
# columnTitles, decoratedResultsDict = \
# gsPerTrackResultsModel.generateColumnTitlesAndResultsDict(choices.leadAttribute)
# else:
# columnTitles, decoratedResultsDict = \
# gsPerTrackResultsModel.generateColumnTitlesAndResultsDict()
#
# core = HtmlCore()
# core.begin()
# core.divBegin(divId='results-page')
# core.divBegin(divClass='results-section')
# core.header(analysisQuestion)
# topTrackTitle = results.keys()[0]
# core.paragraph('''
# The track "%s" is the most atypical track of the GSuite with %s %s similarity to the rest of the tracks
# as measured by the "%s" track similarity measure.
# ''' % (topTrackTitle, strWithNatLangFormatting(results[topTrackTitle]), summaryFunc, similaryStatClassName))
# # core.tableFromDictionary(results, columnNames=['Track title', 'Similarity to rest of tracks in suite (' + summaryFunc+')'], sortable=False)
#
# from quick.util import CommonFunctions
# rawDataURIList = CommonFunctions.getHyperlinksForRawTableData(
# dataDict=decoratedResultsDict, colNames=columnTitles,
# tableId="resultsTable", galaxyFn=galaxyFn)
# core.tableFromDictionary(decoratedResultsDict, columnNames=columnTitles, sortable=True,
# tableId='resultsTable', addInstruction=True,
# addRawDataSelectBox=True, rawDataURIList=rawDataURIList)
# # core.tableFromDictionary(decoratedResultsDict, columnNames=columnTitles, sortable=True, tableId='resultsTable')
#
# columnInd = 0
# if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
# columnInd = 1
# res = GSuiteTracksCoincidingWithQueryTrackTool.drawPlot(
# results, additionalResultsDict,
# 'Similarity to rest of tracks in suite (%s)' % summaryFunc,
# columnInd=columnInd)
# core.line(res)
# core.divEnd()
# core.divEnd()
# core.end()
#
# if choices.addResults == 'Yes':
# GSuiteStatUtils.addResultsToInputGSuite(
# gsuite, results, ['Similarity_score'],
# cls.extraGalaxyFn[GSUITE_EXPANDED_WITH_RESULT_COLUMNS_FILENAME])
elif analysisQuestion == cls.Q3:
q2TS = TrackStructureV2()
randTvProvider = cls.createTrackViewProvider(
choices, ts, analysisBins, genome)
localAnalysis = randTvProvider.supportsLocalAnalysis()
tsRand = getRandomizedVersionOfTs(ts, randTvProvider)
for key in ts.keys():
realTS = TrackStructureV2()
realTS['query'] = ts[key]
realTS['reference'] = FlatTracksTS(
dict([(refKey, refSTS)
for refKey, refSTS in ts.iteritems()
if refKey != key]))
randTS = TrackStructureV2()
randTS['query'] = tsRand[key]
randTS['reference'] = FlatTracksTS([
(refKey, refSTS) for refKey, refSTS in tsRand.iteritems()
if refKey != key
])
hypothesisTS = TrackStructureV2()
hypothesisTS['real'] = realTS
hypothesisTS['rand'] = randTS
q2TS[key] = hypothesisTS
mcfdrDepth = choices.mcfdrDepth if choices.mcfdrDepth else \
AnalysisDefHandler(REPLACE_TEMPLATES['$MCFDRv5$']).getOptionsAsText().values()[0][0]
analysisDefString = REPLACE_TEMPLATES[
'$MCFDRv5$'] + ' -> ' + ' -> MultipleRandomizationManagerStat'
analysisSpec = AnalysisDefHandler(analysisDefString)
analysisSpec.setChoice('MCFDR sampling depth', mcfdrDepth)
analysisSpec.addParameter(
'rawStatistic',
SummarizedInteractionWithOtherTracksV2Stat.__name__)
analysisSpec.addParameter(
'pairwiseStatistic', GSuiteStatUtils.
PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similaryStatClassName])
analysisSpec.addParameter(
'summaryFunc',
GSuiteStatUtils.SUMMARY_FUNCTIONS_MAPPER[summaryFunc])
analysisSpec.addParameter('multitrackSummaryFunc', 'raw')
analysisSpec.addParameter('tail', 'right-tail')
analysisSpec.addParameter('evaluatorFunc',
'evaluatePvalueAndNullDistribution')
analysisSpec.addParameter('runLocalAnalysis',
'Yes' if localAnalysis else 'No')
results = doAnalysis(analysisSpec, analysisBins,
q2TS).getGlobalResult()
resultsTuples = []
for key, res in results['Result'].iteritems():
curRes = res.getResult()
curPval = curRes['P-value']
curTestStat = curRes[
'TSMC_' +
SummarizedInteractionWithOtherTracksV2Stat.__name__]
resultsTuples.append((key, [curTestStat, curPval]))
resultsDict = OrderedDict(
sorted(resultsTuples,
key=lambda t: (-t[1][1], t[1][0]),
reverse=True))
core = HtmlCore()
gsPerTrackResultsModel = GSuitePerTrackResultModel(
resultsDict, [
'Similarity to rest of tracks in suite (%s)' % summaryFunc,
'P-value'
],
additionalResultsDict=additionalResultsDict,
additionalAttributesDict=additionalAttributesDict)
if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
columnTitles, decoratedResultsDict = \
gsPerTrackResultsModel.generateColumnTitlesAndResultsDict(choices.leadAttribute)
else:
columnTitles, decoratedResultsDict = \
gsPerTrackResultsModel.generateColumnTitlesAndResultsDict()
core.begin()
core.divBegin(divId='results-page')
core.divBegin(divClass='results-section')
core.header(analysisQuestion)
topTrackTitle = resultsDict.keys()[0]
core.paragraph('''
The track "%s" has the lowest P-value of %s corresponding to %s %s similarity to the rest of the tracks
as measured by "%s" track similarity measure.
''' % (topTrackTitle,
strWithNatLangFormatting(resultsDict[topTrackTitle][1]),
strWithNatLangFormatting(resultsDict[topTrackTitle][0]),
summaryFunc, similaryStatClassName))
# core.tableFromDictionary(results, columnNames=['Track title', 'Similarity to rest of tracks in suite (' + summaryFunc+')', 'P-value'], sortable=False)
addTableWithTabularAndGsuiteImportButtons(
core,
choices,
galaxyFn,
cls.Q3_SHORT,
decoratedResultsDict,
columnTitles,
gsuite=gsuite,
results=resultsDict,
gsuiteAppendAttrs=['similarity_score', 'p_value'],
sortable=True)
core.divEnd()
core.divEnd()
core.end()
else: # Q4
# mcfdrDepth = choices.mcfdrDepth if choices.mcfdrDepth else \
# AnalysisDefHandler(REPLACE_TEMPLATES['$MCFDR$']).getOptionsAsText().values()[0][0]
# analysisDefString = REPLACE_TEMPLATES['$MCFDRv3$'] + ' -> CollectionSimilarityHypothesisWrapperStat'
# analysisSpec = AnalysisDefHandler(analysisDefString)
# analysisSpec.setChoice('MCFDR sampling depth', mcfdrDepth)
# analysisSpec.addParameter('assumptions', 'PermutedSegsAndIntersegsTrack')
# analysisSpec.addParameter('rawStatistic', 'MultitrackSummarizedInteractionV2Stat')
# analysisSpec.addParameter('pairwiseStatistic', GSuiteStatUtils.PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similaryStatClassName])
# analysisSpec.addParameter('summaryFunc', GSuiteStatUtils.SUMMARY_FUNCTIONS_MAPPER[summaryFunc])
# analysisSpec.addParameter('multitrackSummaryFunc', 'avg') # should it be a choice?
# analysisSpec.addParameter('tail', 'right-tail')
# results = doAnalysis(analysisSpec, analysisBins, tracks).getGlobalResult()
mcfdrDepth = choices.mcfdrDepth if choices.mcfdrDepth else \
AnalysisDefHandler(REPLACE_TEMPLATES['$MCFDRv4$']).getOptionsAsText().values()[0][0]
analysisDefString = REPLACE_TEMPLATES[
'$MCFDRv4$'] + ' -> RandomizationManagerV3Stat'
analysisSpec = AnalysisDefHandler(analysisDefString)
analysisSpec.setChoice('MCFDR sampling depth', mcfdrDepth)
analysisSpec.addParameter('rawStatistic',
'MultitrackSummarizedInteractionV2Stat')
analysisSpec.addParameter(
'pairwiseStatistic', GSuiteStatUtils.
PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similaryStatClassName]
) # needed for call of non randomized stat for assertion
analysisSpec.addParameter(
'summaryFunc',
GSuiteStatUtils.SUMMARY_FUNCTIONS_MAPPER[summaryFunc])
analysisSpec.addParameter('multitrackSummaryFunc',
'avg') # should it be a choice?
analysisSpec.addParameter('tail', 'right-tail')
analysisSpec.addParameter(
'tvProviderClass',
getClassName(
createTrackViewProvider(choices.randType,
choices.randAlg)))
results = doAnalysis(analysisSpec, analysisBins,
ts).getGlobalResult()
pval = results['P-value']
observed = results['TSMC_MultitrackSummarizedInteractionV2Stat']
significanceLevel = 'strong' if pval < 0.01 else (
'weak' if pval < 0.05 else 'no')
core = HtmlCore()
core.begin()
core.divBegin(divId='results-page')
core.divBegin(divClass='results-section')
core.header(analysisQuestion)
core.paragraph('''
The tracks in the suite show %s significance in their collective similarity
(average similarity of a track to the rest) of %s
and corresponding p-value of %s,
as measured by "%s" track similarity measure.
''' % (significanceLevel, strWithNatLangFormatting(observed),
strWithNatLangFormatting(pval), similaryStatClassName))
core.divEnd()
core.divEnd()
core.end()
print str(core)
class MockTrack(Track):
def __new__(cls, *args):
return object.__new__(cls)
def __init__(self, name):
object.__init__(self)
self.trackName = [
name
] #because needed by hash in makeTreeSegregatedByCategory
self.trackTitle = name
q = SingleTrackTS(MockTrack('t1'), OrderedDict(title='t1', cat='q'))
flat = FlatTracksTS()
flat['t2'] = SingleTrackTS(MockTrack('t2'), OrderedDict(title='t2', cat='C1'))
flat['t3'] = SingleTrackTS(MockTrack('t3'), OrderedDict(title='t3', cat='C1'))
flat['t4'] = SingleTrackTS(MockTrack('t4'), OrderedDict(title='t4', cat='C2'))
r = flat.getSplittedByCategoryTS('cat')
orig = TrackStructureV2()
orig['q'] = q
orig['r'] = r
def allInOne():
print 'Orig: ', orig
rerooted = orig.makeTreeSegregatedByCategory(orig['r'])
print 'Rerooted: ', rerooted