def _compute(self):
kwArgs = copy(self._kwArgs)
if 'rawStatistic' in kwArgs:
del kwArgs['rawStatistic']
matrixElRes = []
tr1Subtypes = ProcTrackOptions.getSubtypes(self.getGenome(),
self._track.trackName, True)
tr2Subtypes = ProcTrackOptions.getSubtypes(self.getGenome(),
self._track2.trackName,
True)
assert len(tr1Subtypes) > 0, str(self._track.trackName)
assert len(tr2Subtypes) > 0, str(self._track2.trackName)
if 'minimal' in self._kwArgs:
tr1Subtypes = tr1Subtypes[:1]
tr2Subtypes = tr2Subtypes[:1]
for subtype1 in tr1Subtypes: #['0','1']:
#for subtype2 in ['0','1']:
for subtype2 in tr2Subtypes:
# print ','
tn1 = self._track.trackName + [subtype1]
tn2 = self._track2.trackName + [subtype2]
if not os.path.exists(createDirPath(
tn1, self.getGenome())) or not os.path.exists(
createDirPath(tn2, self.getGenome())):
raise IncompatibleTracksError
#print ','
track1 = Track(tn1)
track1.formatConverters = self._track.formatConverters
track2 = Track(tn2)
track2.formatConverters = self._track2.formatConverters
#self._addChild(self._rawStatistic(self._region, track1, track2, **kwArgs) )
matrixElRes.append(
self._rawStatistic(self._region, track1, track2,
**kwArgs).getResult())
ResultsMemoizer.flushStoredResults()
#assert len(self._children) == 7
#return dict(zip( '00,01,10,11'.split(','), [x.getResult() for x in self._children[3:]]))
allChildRes = array(matrixElRes)
#allChildRes = array([x.getResult() for x in self._children[3:]])
allChildRes = allChildRes.reshape((len(tr1Subtypes), len(tr2Subtypes)))
return {
'Result':
OrderedDict([('Matrix', allChildRes.tolist()),
('Rows', tr1Subtypes), ('Cols', tr2Subtypes)])
}
def _compute(self):
kwArgs = copy(self._kwArgs)
if 'rawStatistic' in kwArgs:
del kwArgs['rawStatistic']
matrixElRes = []
tr1Subtypes = ProcTrackOptions.getSubtypes(self.getGenome(), self._track.trackName, True)
assert len(tr1Subtypes) > 0
for subtype1 in tr1Subtypes:#['0','1']:
for subtype2 in ['0','1']:
tn1 = self._track.trackName + [subtype1]
tn2 = self._track2.trackName + [subtype2]
if not os.path.exists(createDirPath(tn1,self.getGenome())) or not os.path.exists(createDirPath(tn2,self.getGenome())):
raise IncompatibleTracksError
#print ','
track1 = Track( tn1)
track1.formatConverters = self._track.formatConverters
track2 = Track( tn2)
track2.formatConverters = self._track2.formatConverters
#self._addChild(self._rawStatistic(self._region, track1, track2, **kwArgs) )
matrixElRes.append( self._rawStatistic(self._region, track1, track2, **kwArgs).getResult() )
ResultsMemoizer.flushStoredResults()
#assert len(self._children) == 7
#return dict(zip( '00,01,10,11'.split(','), [x.getResult() for x in self._children[3:]]))
allChildRes = array(matrixElRes)
#allChildRes = array([x.getResult() for x in self._children[3:]])
allChildRes = allChildRes.reshape((-1,2))
return OrderedDict([('Matrix', allChildRes.tolist()), ('Rows', tr1Subtypes), ('Cols', ['Case','Control'])])
def getOptionsBox15(prevChoices):
if prevChoices[-2] == 'Select single reference track':
assert prevChoices[4] == 'DHS'
genome = prevChoices[0]
tn = 'Private:GK:Psych:DHSs'.split(':')
return ProcTrackOptions.getSubtypes(genome, tn, True)
elif prevChoices[-2] == 'Select a range among all reference tracks':
return ''
def getOptionsBoxTfTracks(cls, prevChoices):
if prevChoices.sourceTfsDetails != cls.SELECT:
genome = prevChoices.genome
sourceTfs = prevChoices.sourceTfs
sourceTfsDetails = prevChoices.sourceTfsDetails
if sourceTfs == cls.SELECT:
return
elif sourceTfs == 'Hyperbrowser repository':
tfSourceTN = TfTrackNameMappings.getTfTrackNameMappings(
prevChoices.genome)[sourceTfsDetails]
subtypes = ProcTrackOptions.getSubtypes(
prevChoices.genome, tfSourceTN, True)
falses = ['False'] * len(subtypes)
return OrderedDict(zip(subtypes, falses))
elif sourceTfs == cls.REGIONS_FROM_HISTORY:
if isinstance(sourceTfsDetails, basestring):
galaxyTN = sourceTfsDetails.split(':')
if galaxyTN[
1] == "gsuite": #ExternalTrackManager.extractFileSuffixFromGalaxyTN(prevChoices.sourceTfsDetails, allowUnsupportedSuffixes=True) == "gsuite"
errorString = GeneralGuiTool._checkGSuiteFile(
sourceTfsDetails)
if not errorString:
gSuite = getGSuiteFromGalaxyTN(sourceTfsDetails)
sizeErrorString = GeneralGuiTool._checkGSuiteTrackListSize(
gSuite, 1, 1000)
if not sizeErrorString:
reqErrorString = GeneralGuiTool._checkGSuiteRequirements \
(gSuite,
AllTfsOfRegions.GSUITE_ALLOWED_FILE_FORMATS,
AllTfsOfRegions.GSUITE_ALLOWED_LOCATIONS,
AllTfsOfRegions.GSUITE_ALLOWED_TRACK_TYPES,
AllTfsOfRegions.GSUITE_DISALLOWED_GENOMES)
if not reqErrorString:
validity = 'Valid'
else:
return
else:
return
else:
return
if validity == 'Valid':
selectedTrackNames = []
gSuite = getGSuiteFromGalaxyTN(sourceTfsDetails)
for track in gSuite.allTracks():
selectedTrackNames.append(':'.join(
track.trackName))
falses = ['False'] * len(selectedTrackNames)
return OrderedDict(zip(selectedTrackNames, falses))
else:
tfTrackName = ExternalTrackManager.getPreProcessedTrackFromGalaxyTN(
genome, galaxyTN)
return [':'.join(tfTrackName)]
else:
return
else:
return
else:
return
def getOptionsBoxDataset(prevChoices):
#if prevChoices.genome == 'hg19':
if prevChoices.genome in ['hg19', 'mm9']:
parentTrack = ColocalizationIn3DTool.PARENT_TRACKNAME + [
prevChoices.interactions
] + [prevChoices.cellLine]
return ProcTrackOptions.getSubtypes(prevChoices.genome,
parentTrack,
fullAccess=False)
def _getAllDiseases(prevChoices):
if isinstance(prevChoices[2], dict):
return prevChoices[2].keys()
else:
tn = SelectDiseaseTool._getDiseaseTn(prevChoices)
return [
x for x in ProcTrackOptions.getSubtypes(
SelectDiseaseTool.GENOME, tn)
]
def _getAllPwms(prevChoices):
if isinstance(prevChoices[2], dict):
return prevChoices[2].keys()
else:
tfTrackName = SelectTfTool.TRACK_DICT[prevChoices[0]]
return [
pwm for pwm in ProcTrackOptions.getSubtypes(
SelectTfTool.GENOME, tfTrackName)
]
def getOptionsBox3(prevChoices):
seedSource = prevChoices[1]
if seedSource == 'TFBS from history':
return ('__history__', 'bed')
else:
genome = prevChoices[0]
tfSourceTN = TfInfo.getTfTrackNameMappings(genome)[prevChoices[1]]
subtypes = ProcTrackOptions.getSubtypes(genome, tfSourceTN, True)
return subtypes
def getOptionsBox3(prevChoices):
seedSource = prevChoices[1]
if seedSource == 'TFBS from history':
return ('__history__','bed')
else:
genome = prevChoices[0]
tfSourceTN = TfInfo.getTfTrackNameMappings(genome)[ prevChoices[1] ]
subtypes = ProcTrackOptions.getSubtypes(genome, tfSourceTN, True)
return subtypes
def superLine2batch(cls, line, genome):
if line.strip() == '' or line[0] == '#':
return []
cols = line.split(BATCH_COL_SEPARATOR)
if len(cols) == 4:
cols = cols[:3] + ['dummy', cols[3]]
line = BATCH_COL_SEPARATOR.join(cols)
from quick.batch.BatchRunner import BatchRunner
#errorResult, userBinSource = BatchRunner._constructBins(cols[0], cols[1], genome)
#if errorResult is not None:
# return [BATCH_COL_SEPARATOR.join([ 'dummy', cols[0], cols[1], 'dummyTN1', 'dummyTN2', 'dummyStatName' ])]
for colIndex, col in zip(range(
2, 5), cols[2:5]): #not binSpec, but tn1, tn2, statistic..
if '/' in col:
splitPoint = col.find('/')
splittedCols = col[0:splitPoint], col[splitPoint + 1:]
splittedListLines = [ cols[0:colIndex] +[splitCol]+ cols[colIndex+1:]\
for splitCol in splittedCols]
#1+''
return reduce(lambda x, y: x + y, [
cls.superLine2batch(BATCH_COL_SEPARATOR.join(line), genome)
for line in splittedListLines
])
for colIndex, col in zip(range(2, 4), cols[2:4]):
if '*' in col:
typeParts = col.split(':')
assert typeParts[-1] == '*' and not any(
'*' in part for part in typeParts[:-1])
#whitespacedTrackName = [x.replace('_',' ') for x in typeParts[:-1]]
unquotedTrackName = [unquote(x) for x in typeParts[:-1]]
#print 'TEMP: ',genome, unquotedTrackName
starOptions = ProcTrackOptions.getSubtypes(
genome, unquotedTrackName, True)
if len(starOptions) == 0:
raise Exception('No subtracks for parent track: %s' %
str(unquotedTrackName))
#typeOptions = [':'.join(typeParts[:-1] + [starOpt.replace(' ','_')]) for starOpt in starOptions]
typeOptions = [
':'.join(typeParts[:-1] + [quote(starOpt)])
for starOpt in starOptions
]
splittedCols = typeOptions
splittedListLines = [ cols[0:colIndex] +[splitCol]+ cols[colIndex+1:]\
for splitCol in splittedCols]
return reduce(lambda x, y: x + y, [
cls.superLine2batch(BATCH_COL_SEPARATOR.join(line), genome)
for line in splittedListLines
])
return [cls.generateBatchName(line) + BATCH_COL_SEPARATOR + line]
def getOptionsBox6(cls, prevChoices):
if prevChoices[4] == cls.REGIONS_FROM_HISTORY:
return ('__history__','bed','bedgraph')
else:
tfSourceTN = TfInfo.getTfTrackNameMappings(prevChoices[1])[ prevChoices[4] ]
genome = prevChoices[1]
subtypes = ProcTrackOptions.getSubtypes(genome, tfSourceTN, True)
#logMessage(str(subtypes))
#return ['V$AHR_01']
return subtypes
def getOptionsBox6(cls, prevChoices):
if prevChoices[4] == cls.REGIONS_FROM_HISTORY:
return ('__history__', 'bed', 'bedgraph')
else:
tfSourceTN = TfInfo.getTfTrackNameMappings(
prevChoices[1])[prevChoices[4]]
genome = prevChoices[1]
subtypes = ProcTrackOptions.getSubtypes(genome, tfSourceTN, True)
#logMessage(str(subtypes))
#return ['V$AHR_01']
return subtypes
def _calcAndStoreSubTrackCount(self, trackName):
ti = TrackInfo(self._genome, trackName)
trackCount = 0
for subTrackName in ProcTrackOptions.getSubtypes(self._genome, trackName, True):
subTrackCount = TrackInfo(self._genome, trackName + [subTrackName]).subTrackCount
if subTrackCount:
trackCount += subTrackCount
if ti.isValid():
trackCount += 1
ti.subTrackCount = trackCount
ti.store()
def getOptionsBoxTrack2(cls, prevChoices):
if prevChoices[4] == 'yes':
genome = prevChoices[1]
prefixTN = cls.STD_PREFIX_TN
trackList = ProcTrackOptions.getSubtypes(genome, prefixTN, True)
if cls._cacheDict.get('track2') in trackList:
trackList.remove(cls._cacheDict.get('track2'))
trackList.insert(0, cls._cacheDict.get('track2'))
return [cls.NO_TRACK_SHORTNAME] + trackList
else:
return '__track__', 'history'
def getOptionsBoxTrack2(cls, prevChoices):
if prevChoices[4]=='yes' :
genome = prevChoices[1]
prefixTN = cls.STD_PREFIX_TN
trackList = ProcTrackOptions.getSubtypes(genome, prefixTN, True)
if cls._cacheDict.get('track2') in trackList:
trackList.remove(cls._cacheDict.get('track2'))
trackList.insert(0, cls._cacheDict.get('track2'))
return [cls.NO_TRACK_SHORTNAME] + trackList
else:
return '__track__','history'
def _calcAndStoreSubTrackCount(self, trackName):
ti = TrackInfo(self._genome, trackName)
trackCount = 0
for subTrackName in ProcTrackOptions.getSubtypes(
self._genome, trackName, True):
subTrackCount = TrackInfo(self._genome,
trackName + [subTrackName]).subTrackCount
if subTrackCount:
trackCount += subTrackCount
if ti.isValid():
trackCount += 1
ti.subTrackCount = trackCount
ti.store()
def getOptionsBoxTfTracks(cls, prevChoices):
if prevChoices.sourceTfs:
if prevChoices.sourceTfs == cls.REGIONS_FROM_HISTORY:
return ('__history__', 'bed', 'category.bed', 'gtrack')
elif prevChoices.sourceTfs == cls.SELECT:
return
else:
tfSourceTN = TfTrackNameMappings.getTfTrackNameMappings(
prevChoices.genome)[prevChoices.sourceTfs]
subtypes = ProcTrackOptions.getSubtypes(
prevChoices.genome, tfSourceTN, True)
return subtypes
else:
return
def getOptionsBox6(cls, prevChoices):
#if prevChoices[0]:
# cls.updateCacheDict(prevChoices[0])
if prevChoices[4] =='yes' :
genome = prevChoices[3]
prefixTN = cls.STD_PREFIX_TN
trackList = ProcTrackOptions.getSubtypes(genome, prefixTN, True)
#if cls._cacheDict.get('track1') in trackList:
# trackList.remove(cls._cacheDict.get('track1'))
# trackList.insert(0, cls._cacheDict.get('track1'))
return trackList
else:
return '__track__','history'
def getOptionsBox6(cls, prevChoices):
#if prevChoices[0]:
# cls.updateCacheDict(prevChoices[0])
if prevChoices[4] == 'yes':
genome = prevChoices[3]
prefixTN = cls.STD_PREFIX_TN
trackList = ProcTrackOptions.getSubtypes(genome, prefixTN, True)
#if cls._cacheDict.get('track1') in trackList:
# trackList.remove(cls._cacheDict.get('track1'))
# trackList.insert(0, cls._cacheDict.get('track1'))
return trackList
else:
return '__track__', 'history'
def getOptionsBoxInteractions(
prevChoices): # Alternatively: getOptionsBoxKey2()
'''
See getOptionsBox1().
prevChoices is a namedtuple of selections made by the user in the
previous input boxes (that is, a namedtuple containing only one element
in this case). The elements can accessed either by index, e.g.
prevChoices[0] for the result of input box 1, or by key, e.g.
prevChoices.key (case 2).
'''
#if prevChoices.genome == 'hg19':
if prevChoices.genome in ['hg19', 'mm9']:
parentTrack = ColocalizationIn3DTool.PARENT_TRACKNAME
return ProcTrackOptions.getSubtypes(prevChoices.genome,
parentTrack,
fullAccess=False)
def superLine2batch(cls, line, genome):
if line.strip() == '' or line[0]=='#':
return []
cols = line.split(BATCH_COL_SEPARATOR)
if len(cols) == 4:
cols = cols[:3] + ['dummy', cols[3]]
line = BATCH_COL_SEPARATOR.join(cols)
from quick.batch.BatchRunner import BatchRunner
#errorResult, userBinSource = BatchRunner._constructBins(cols[0], cols[1], genome)
#if errorResult is not None:
# return [BATCH_COL_SEPARATOR.join([ 'dummy', cols[0], cols[1], 'dummyTN1', 'dummyTN2', 'dummyStatName' ])]
for colIndex, col in zip(range(2,5), cols[2:5]):
if '/' in col:
splitPoint = col.find('/')
splittedCols = col[0:splitPoint], col[splitPoint+1:]
splittedListLines = [ cols[0:colIndex] +[splitCol]+ cols[colIndex+1:]\
for splitCol in splittedCols]
#1+''
return reduce(lambda x,y:x+y, [cls.superLine2batch(BATCH_COL_SEPARATOR.join(line), genome) for line in splittedListLines])
for colIndex, col in zip(range(2,4), cols[2:4]):
if '*' in col:
typeParts = col.split(':')
assert typeParts[-1]=='*' and not any('*' in part for part in typeParts[:-1])
#whitespacedTrackName = [x.replace('_',' ') for x in typeParts[:-1]]
unquotedTrackName = [unquote(x) for x in typeParts[:-1]]
print 'TEMP: ',genome, unquotedTrackName
starOptions = ProcTrackOptions.getSubtypes(genome, unquotedTrackName, True)
if len(starOptions) == 0:
raise Exception('No subtracks for parent track: %s' % str(unquotedTrackName))
#typeOptions = [':'.join(typeParts[:-1] + [starOpt.replace(' ','_')]) for starOpt in starOptions]
typeOptions = [':'.join(typeParts[:-1] + [starOpt]) for starOpt in starOptions]
splittedCols = typeOptions
splittedListLines = [ cols[0:colIndex] +[splitCol]+ cols[colIndex+1:]\
for splitCol in splittedCols]
return reduce(lambda x,y:x+y, [cls.superLine2batch(BATCH_COL_SEPARATOR.join(line), genome) for line in splittedListLines])
return [cls.generateBatchName(line)+BATCH_COL_SEPARATOR+line]
def yielder(self, curTn):
if self._avoidLiterature and curTn == GenomeInfo.getPropertyTrackName(self._genome, 'literature'):
return
for subtype in ProcTrackOptions.getSubtypes(self._genome, curTn, self._fullAccess):
#if self._avoidLiterature and subtype == 'Literature':
if subtype[0] in ['.','_']:
continue
newTn = curTn + [subtype]
doBreak = False
for subTn in self.yielder(newTn):
yield subTn
if ProcTrackOptions.isValidTrack(self._genome, curTn, self._fullAccess):
yield curTn
def yielder(self, curTn, level=0):
if self._avoidLiterature and curTn == GenomeInfo.getPropertyTrackName(
self._genome, 'literature'):
return
for subtype in ProcTrackOptions.getSubtypes(self._genome, curTn,
self._fullAccess):
#if self._avoidLiterature and subtype == 'Literature':
if subtype[0] in ['.', '_']:
continue
newTn = curTn + [subtype]
doBreak = False
for subTn in self.yielder(newTn, level=level + 1):
yield subTn
if self._includeParentTrack or level > 0:
if ProcTrackOptions.isValidTrack(self._genome, curTn,
self._fullAccess):
yield curTn
print 'FAILED: Error when preprocessing TestGenome. Error:'
print ' ' + str(e).strip()
sys.exit(1)
for allowOverlaps in [False, True]:
fromDir = createDirPath(['GESourceTracks'], 'TestGenome', allowOverlaps=allowOverlaps)
toDir = createDirPath([], 'ModelsForExternalTracks', allowOverlaps=allowOverlaps)
try:
if not os.path.exists(toDir):
shutil.copytree(fromDir, toDir)
print 'OK: Copied from %s to %s.' % (fromDir, toDir)
except Exception, e:
print 'FAILED: Error occurred copying from %s to %s: ' % (fromDir, toDir) + str(e).strip()
sys.exit(1)
for track in ProcTrackOptions.getSubtypes('TestGenome', ['GESourceTracks']):
ti = TrackInfo('TestGenome', ['GESourceTracks', track])
ti.trackName = [track]
ti.genome = 'ModelsForExternalTracks'
ti.store()
from quick.util.GenomeInfo import GenomeInfo
from datetime import datetime
gi = GenomeInfo('TestGenome')
gi.fullName = 'TestGenome'
gi.sourceUrls = ['http://hgdownload.cse.ucsc.edu/goldenPath/hg18/chromosomes/chr21.fa.gz', \
'http://hgdownload.cse.ucsc.edu/goldenPath/hg18/chromosomes/chrM.fa.gz']
gi.sourceChrNames = ['chr21', 'chrM']
gi.installedBy = 'Setup.py'
gi.genomeBuildSource = 'NCBI'
gi.genomeBuildName = 'hg18'
fromDir = createDirPath(['GESourceTracks'],
'TestGenome',
allowOverlaps=allowOverlaps)
toDir = createDirPath([],
'ModelsForExternalTracks',
allowOverlaps=allowOverlaps)
try:
if not os.path.exists(toDir):
shutil.copytree(fromDir, toDir)
print 'OK: Copied from %s to %s.' % (fromDir, toDir)
except Exception, e:
print 'FAILED: Error occurred copying from %s to %s: ' % (
fromDir, toDir) + str(e).strip()
sys.exit(1)
for track in ProcTrackOptions.getSubtypes('TestGenome',
['GESourceTracks']):
ti = TrackInfo('TestGenome', ['GESourceTracks', track])
ti.trackName = [track]
ti.genome = 'ModelsForExternalTracks'
ti.store()
from quick.util.GenomeInfo import GenomeInfo
from datetime import datetime
gi = GenomeInfo('TestGenome')
gi.fullName = 'TestGenome'
gi.sourceUrls = ['http://hgdownload.cse.ucsc.edu/goldenPath/hg18/chromosomes/chr21.fa.gz', \
'http://hgdownload.cse.ucsc.edu/goldenPath/hg18/chromosomes/chrM.fa.gz']
#gi.sourceChrNames = ['chr21', 'chrM']
gi.installedBy = 'Setup.py'
gi.genomeBuildSource = 'NCBI'
gi.genomeBuildName = 'hg18'
def _getAllDiseases(prevChoices):
if isinstance(prevChoices[2], dict):
return prevChoices[2].keys()
else:
tn = SelectDiseaseTool._getDiseaseTn(prevChoices)
return [x for x in ProcTrackOptions.getSubtypes(SelectDiseaseTool.GENOME, tn)]
def _getAllPwms(prevChoices):
if isinstance(prevChoices[2], dict):
return prevChoices[2].keys()
else:
tfTrackName = SelectTfTool.TRACK_DICT[prevChoices[0]]
return [pwm for pwm in ProcTrackOptions.getSubtypes(SelectTfTool.GENOME, tfTrackName)]