def generateQ1output(cls, additionalResultsDict, analysisQuestion, choices, galaxyFn,
gsPerTrackResults, queryTrackTitle, gsuite, results,
similarityStatClassName):
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" in the GSuite is the one most similar to the query track %s, with a similarity score of %s
as measured by the "%s" track similarity measure.
''' % (
topTrackTitle, queryTrackTitle, strWithNatLangFormatting(results[topTrackTitle]),
similarityStatClassName))
core.divBegin()
addTableWithTabularAndGsuiteImportButtons(
core, choices, galaxyFn, cls.Q1_SHORT, tableDict=gsPerTrackResults[1],
columnNames=gsPerTrackResults[0], gsuite=gsuite, results=results,
gsuiteAppendAttrs=['similarity_score'], sortable=True)
core.divEnd()
columnInd = 0
if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
columnInd = 1
res = GSuiteTracksCoincidingWithQueryTrackTool.drawPlot(
results, additionalResultsDict,
'Similarity to query track', columnInd=columnInd)
core.line(res)
core.divEnd()
core.divEnd()
core.end()
return core
def execute(cls, choices, galaxyFn=None, username=''):
cls._setDebugModeIfSelected(choices)
gSuite = getGSuiteFromGalaxyTN(choices.gsuite)
if gSuite.genome != choices.genome:
gSuite.setGenomeOfAllTracks(choices.genome)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
paragraphs = []
paragraphs += generatePilotPageTwoParagraphs(gSuite,
galaxyFn,
regSpec=regSpec,
binSpec=binSpec)
paragraphs += generatePilotPageThreeParagraphs(gSuite,
galaxyFn,
regSpec=regSpec,
binSpec=binSpec)
core = HtmlCore()
core.begin()
core.divBegin(divId='results-page')
core.divBegin(divClass='results-section')
core.header('Similarity and uniqueness of tracks')
for prg in paragraphs:
core.paragraph(prg)
core.divEnd()
core.divEnd()
core.end()
print core
def createDescription(toolDescription=None, stepsToRunTool=None, toolResult=None, limitation=None):
core = HtmlCore()
if toolDescription!=None or stepsToRunTool!=None or toolResult!=None or limitation!=None:
core.divBegin(divId='decription-page')
core.divBegin(divClass='decription-section')
core.header('Description')
#small description of tool (The resaon of creating the tool)
if toolDescription!=None:
core.divBegin(divClass='decription-section-main')
core.paragraph(toolDescription)
core.divEnd()
#how to use tool
if stepsToRunTool!=None:
core.paragraph('To run the tool, follow these steps:')
core.orderedList(stepsToRunTool)
#what is the result of tool
if toolDescription!=None:
core.divBegin(divClass='decription-section-main')
core.paragraph(toolResult)
core.divEnd()
#what are the limitation for tool
# if limitation:
# limits...
core.divEnd()
core.divEnd()
return str(core)
def execute(cls, choices, galaxyFn=None, username=''):
#rsids = choices.rsid.split()
if choices.run == 'Batch':
#print rsids
return cls.execute_batch(choices, galaxyFn, username)
elif choices.batch != '__batch__':
print '<div class="debug">'
results = GalaxyRunSpecificFile(['html'], galaxyFn)
dir = os.path.dirname(results.getDiskPath(ensurePath=True))
os.mkdir(dir + '/html')
#print '<div class="debug">'
cls.choices = choices
cls.run_varmelt(dir, choices)
url = results.getURL()
if choices.run == 'Single' and choices.batch != '__batch__':
print '</div></pre>'
core = HtmlCore()
core.header('Primer3 candidates')
VariantMeltingProfile.primer3_table_header(core)
VariantMeltingProfile.primer3_resultsfile_header(dir)
for r in range(0, int(choices.numReturn)):
datafile = dir + '/tempdata.' + str(r) + '.results.txt'
if os.path.exists(datafile):
variant_pos = VariantMeltingProfile.proc_temp_data(dir, str(r))
chart = open(dir + '/html/chart-' + str(r) + '.html', 'w')
chart.write(VariantMeltingProfile.make_chart(variant_pos, r))
chart.write(cls.primer3_results_table(dir, r))
chart.write('</body></html>')
chart.close()
cls.primer3_results(dir, r)
cls.primer3_results_table(dir, r, core, url)
print '<a href="%s/chart-%d.html">Results/graph num %d</a><br>' % (
url, r, r + 1)
else:
cls.primer3_results(dir, r)
cls.primer3_results_table(dir, r, core, None)
break
core.tableFooter()
if choices.run == 'Single' and choices.batch != '__batch__':
print str(core)
print '<pre>'
xcore = HtmlCore()
xcore.begin()
xcore.append(str(core))
xcore.end()
open(dir + '/results.html', 'w').write(str(xcore))
def execute(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.
'''
resultsFN = ExternalTrackManager.extractFnFromGalaxyTN(
choices.resultsFile)
examResults = TaskScoreOverview(resultsFN, galaxyFn)
examResults.run()
core = HtmlCore()
core.begin()
core.header('Overview of exam scores')
for table in examResults.getTables():
core.divBegin(divClass='resultsTable')
core.tableHeader([])
for key, val in table.iteritems():
core.tableLine([key, val])
core.tableFooter()
core.divEnd()
for plotUrl in examResults.getPlotUrls():
core.divBegin(divClass='plot')
core.image(plotUrl)
core.divEnd()
core.end()
print core
def generateQ2Output(cls, additionalAttributesDict, additionalResultsDict, analysisQuestion, choices,
galaxyFn, queryTrackTitle, gsuite, results, similarityStatClassName):
gsPerTrackResultsModel = GSuitePerTrackResultModel(results, ['Similarity to query track', 'P-value'],
additionalResultsDict=additionalResultsDict,
additionalAttributesDict=additionalAttributesDict)
if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
gsPerTrackResults = gsPerTrackResultsModel.generateColumnTitlesAndResultsDict(choices.leadAttribute)
else:
gsPerTrackResults = 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" has the lowest P-value of %s corresponding to %s similarity to the query track "%s"
as measured by "%s" track similarity measure.
''' % (topTrackTitle, strWithNatLangFormatting(results[topTrackTitle][1]),
strWithNatLangFormatting(results[topTrackTitle][0]), queryTrackTitle, similarityStatClassName))
addTableWithTabularAndGsuiteImportButtons(
core, choices, galaxyFn, cls.Q2_SHORT, tableDict=gsPerTrackResults[1],
columnNames=gsPerTrackResults[0], gsuite=gsuite, results=results,
gsuiteAppendAttrs=['similarity_score', 'p_value'], sortable=True)
columnInd = 0
if choices.leadAttribute and choices.leadAttribute != GSuiteConstants.TITLE_COL:
columnInd = 1
resultsSeparateListPart = OrderedDict()
additionalResultsDictIncludePartFromResults = OrderedDict()
for k, v in results.iteritems():
if k not in resultsSeparateListPart.keys():
resultsSeparateListPart[k] = v[0]
if k not in additionalResultsDictIncludePartFromResults.keys():
additionalResultsDictIncludePartFromResults[k] = OrderedDict()
additionalResultsDictIncludePartFromResults[k]['P-Value'] = v[1]
for k1, v1 in additionalResultsDict[k].iteritems():
additionalResultsDictIncludePartFromResults[k][k1] = v1
res = GSuiteTracksCoincidingWithQueryTrackTool.drawPlot(
resultsSeparateListPart, additionalResultsDictIncludePartFromResults,
'Similarity to query track', columnInd=columnInd)
core.line(res)
core.divEnd()
core.divEnd()
core.end()
return core
def primer3_results_table(cls, dir, rnr, core=None, url='.'):
append = True
if not core:
append = False
core = HtmlCore()
core.header('Primer3 results')
VariantMeltingProfile.primer3_table_header(core)
rows = []
try:
p3 = open(dir + '/primer3.' + str(rnr) + '.results.txt', 'r')
line = p3.readline()
if line:
if url != None:
link = '<a href="%s/chart-%d.html">%d (view)</a>' % (
url, rnr, rnr + 1)
else:
link = str(rnr + 1)
row = [link]
cols = line.strip().split('\t')
for col in cols:
if len(col) > 40:
row.append('<br>'.join(
[col[c:c + 40] for c in xrange(0, len(col), 40)]))
else:
row.append(col)
row += [''] * (len(cls.primer3_headers) - len(cols) - 1)
else:
#XXX
row = [link] + ['?'] * 10 + [line.split()[10]] + ['?']
rows.append(row)
p3.close()
except IOError:
rows.append(
[str(rnr + 1), cls.choices.chr[3:], 'No primers found'] +
['?'] * 8)
for row in rows:
core.tableLine(row)
if not append:
core.tableFooter()
#core.append('<p><a href="javascript:window.history.back()">Go back</a></p>')
return str(core)
def execute(choices, galaxyFn=None, username=''):
gSuite = getGSuiteFromGalaxyTN(choices.gsuite)
if gSuite.genome != choices.genome:
gSuite.setGenomeOfAllTracks(choices.genome)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
paragraphs = generatePilotPageTwoParagraphs(gSuite,
galaxyFn,
regSpec=regSpec,
binSpec=binSpec)
core = HtmlCore()
core.begin()
core.header('Overlap between tracks')
for prg in paragraphs:
core.paragraph(prg)
core.end()
print core
def execute(choices, galaxyFn=None, username=''):
gSuite = getGSuiteFromGalaxyTN(choices.gsuite)
if gSuite.genome != choices.genome:
gSuite.setGenomeOfAllTracks(choices.genome)
# regSpec, binSpec = UserBinSelector.getRegsAndBinsSpec(choices)
paragraphs = generatePilotPageFiveParagraphs(gSuite, galaxyFn)
core = HtmlCore()
core.begin()
core.divBegin(divId='results-page')
core.divBegin(divClass='results-section')
core.header('Clustering of track elements')
for prg in paragraphs:
core.paragraph(prg)
core.divEnd()
core.divEnd()
core.end()
print core
def generateQ3output(cls, analysisQuestion, queryTrackTitle, results, similarityStatClassName):
pval = results['P-value']
observed = results['TSMC_SummarizedInteractionWithOtherTracksV2Stat']
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 query track %s shows %s significance in similarity to the suite of %s
and corresponding p-value of %s,
as measured by "%s" track similarity measure.
''' % (
queryTrackTitle, significanceLevel, strWithNatLangFormatting(observed),
strWithNatLangFormatting(pval),
similarityStatClassName))
core.divEnd()
core.divEnd()
core.end()
return core
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.gtr
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.
'''
fnSource = ExternalTrackManager.extractFnFromGalaxyTN(
choices[2].split(':'))
core = HtmlCore()
core.begin()
valid = False
try:
core.header('Validating GTrack headers')
core.styleInfoBegin(styleClass='debug')
print str(core)
core = HtmlCore()
gtrackSource = GtrackGenomeElementSource(
fnSource,
choices[1] if choices[0] == 'Yes' else None,
printWarnings=True)
core.append('Done')
core.styleInfoEnd()
core.header('Validating complete GTrack file')
core.styleInfoBegin(styleClass='debug')
print str(core)
core = HtmlCore()
try:
for ge in gtrackSource:
pass
except Exception, e:
raise
else:
def _buildContent(self):
#iterate through dictionary and for each key create a section (one of [GSuite, Track...])
# each value in the dictionary is a list of GiudeData objects that go into the section defined by the key
htmlCore = HtmlCore()
htmlCore.divBegin('toolGuideInfo')
htmlCore.divBegin(divClass='toolGuideInfoText')
htmlCore.divBegin(divClass='toolGuideInfoTextHeader')
htmlCore.line(TOOL_GUIDE_HELP_HEADER_TEXT)
htmlCore.divEnd()
htmlCore.divBegin(divClass='toolGuideInfoText')
htmlCore.line(TOOL_GUIDE_HELP_HEADER_TEXT_TEXT)
htmlCore.divEnd()
htmlCore.divEnd()
for guideDataKey, guideDataValues in self._guideDataDict.iteritems():
htmlCore.divBegin('toolGuide')
if guideDataKey in TOOL_INPUT_TYPE_TO_TOOL_GUIDE_HELP_HEADER_DICT:
htmlCore.header(TOOL_INPUT_TYPE_TO_TOOL_GUIDE_HELP_HEADER_DICT[
guideDataKey])
for guideDataValue in guideDataValues:
htmlCore.divBegin(divClass='toolGuideData')
htmlCore.divBegin(divClass='toolGuideImgTitle')
if guideDataValue.imgUrl:
htmlCore.image(guideDataValue.imgUrl)
htmlCore.link(text=guideDataValue.toolDisplayName,
url=str(guideDataValue.toolUrl),
args=(' onclick="%s"' % guideDataValue.onclick)
if guideDataValue.onclick else '')
htmlCore.divEnd()
htmlCore.divBegin(divClass='toolGuideDesc')
htmlCore.append(guideDataValue.description)
if guideDataValue.helpPageUrl:
htmlCore.link(text='...read more',
url=str(guideDataValue.helpPageUrl))
htmlCore.divEnd()
htmlCore.divEnd()
htmlCore.divEnd()
htmlCore.divEnd()
#raise Exception(str(htmlCore))#to debug
self._guideContent = str(htmlCore)
def execute(choices, galaxyFn=None, username=''):
gSuite = getGSuiteFromGalaxyTN(choices.gsuite)
if gSuite.genome != choices.genome:
gSuite.setGenomeOfAllTracks(choices.genome)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
paragraphs = OrderedDict()
paragraphs[
'Basic overview of tracks in collection'] = generatePilotPageOneParagraphs(
gSuite,
galaxyFn,
regSpec=regSpec,
binSpec=binSpec,
username=username)
paragraphs['Overlap between tracks'] = generatePilotPageTwoParagraphs(
gSuite, galaxyFn, regSpec=regSpec, binSpec=binSpec)
paragraphs[
'Similarity and uniqueness of tracks'] = generatePilotPageThreeParagraphs(
gSuite, galaxyFn, regSpec=regSpec, binSpec=binSpec)
paragraphs['Clustering of tracks'] = generatePilotPageFiveParagraphs(
gSuite, galaxyFn)
core = HtmlCore()
core.begin()
core.divBegin(divId='results-page', divClass='trackbook_main')
for hdr, prgList in paragraphs.iteritems():
core.divBegin(divClass='trackbook_section')
core.divBegin(divClass='results-section')
core.header(hdr)
for prg in prgList:
core.paragraph(prg)
core.divEnd()
core.divEnd()
core.divEnd()
core.end()
print core
def execute(cls, choices, galaxyFn=None, username=''):
gSuite = getGSuiteFromGalaxyTN(choices.gsuite)
if gSuite.genome != choices.genome:
gSuite.setGenomeOfAllTracks(choices.genome)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
paragraphs = generatePilotPageOneParagraphs(gSuite,
galaxyFn,
regSpec=regSpec,
binSpec=binSpec,
username=username)
core = HtmlCore()
core.begin()
core.divBegin(divId='results-page')
core.divBegin(divClass='results-section')
core.header('Basic overview of tracks in collection')
for prg in paragraphs:
core.paragraph(prg)
core.divEnd()
core.divEnd()
core.end()
print core
def execute(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.
'''
from gold.application.LogSetup import setupDebugModeAndLogging
setupDebugModeAndLogging()
targetTrackNames, targetTrackCollection, targetTrackGenome = getGSuiteDataFromGalaxyTN(
choices.gSuiteFirst)
targetTracksDict = OrderedDict(
zip(targetTrackNames, targetTrackCollection))
refTrackNames, refTrackCollection, refTrackCollectionGenome = getGSuiteDataFromGalaxyTN(
choices.gSuiteSecond)
assert targetTrackGenome == refTrackCollectionGenome, 'Reference genome must be the same one in both GSuite files.'
refTracksDict = OrderedDict(zip(refTrackNames, refTrackCollection))
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
analysisDef = 'dummy -> RawOverlapStat'
results = OrderedDict()
for targetTrackName, targetTrack in targetTracksDict.iteritems():
for refTrackName, refTrack in refTracksDict.iteritems():
result = GalaxyInterface.runManual([targetTrack, refTrack],
analysisDef,
regSpec,
binSpec,
targetTrackGenome,
galaxyFn,
printRunDescription=False,
printResults=False)
if targetTrackName not in results:
results[targetTrackName] = OrderedDict()
results[targetTrackName][
refTrackName] = result.getGlobalResult()
targetTrackTitles = results.keys()
stat = choices.statistic
statIndex = STAT_LIST_INDEX[stat]
title = stat + ' analysis of track collections'
processedResults = []
headerColumn = []
for targetTrackName in targetTrackTitles:
resultRowDict = processRawResults(results[targetTrackName])
resultColumn = []
headerColumn = []
for refTrackName, statList in resultRowDict.iteritems():
resultColumn.append(statList[statIndex])
headerColumn.append(refTrackName)
processedResults.append(resultColumn)
transposedProcessedResults = [list(x) for x in zip(*processedResults)]
tableHeader = ['Track names'] + targetTrackTitles
htmlCore = HtmlCore()
htmlCore.begin()
htmlCore.header(title)
htmlCore.divBegin('resultsDiv')
htmlCore.tableHeader(tableHeader,
sortable=True,
tableId='resultsTable')
for i, row in enumerate(transposedProcessedResults):
line = [headerColumn[i]] + row
htmlCore.tableLine(line)
htmlCore.tableFooter()
htmlCore.divEnd()
addColumnPlotToHtmlCore(htmlCore,
targetTrackNames,
refTrackNames,
stat,
title + ' plot',
processedResults,
xAxisRotation=315)
htmlCore.hideToggle(styleClass='debug')
htmlCore.end()
print htmlCore
def execute(cls, choices, galaxyFn=None, username=''):
cls._setDebugModeIfSelected(choices)
genome = choices.genome
genomicRegions = choices.genomicRegions
genomicRegionsTracks = choices.genomicRegionsTracks
sourceTfs = choices.sourceTfs
sourceTfsDetails = choices.sourceTfsDetails
tfTracks = choices.tfTracks
# Get Genomic Region track name:
if genomicRegions == cls.REGIONS_FROM_HISTORY:
galaxyTN = genomicRegionsTracks.split(':')
genElementTrackName = ExternalTrackManager.getPreProcessedTrackFromGalaxyTN(
genome, galaxyTN)
#queryGSuite = getGSuiteFromGalaxyTN(genomicRegionsTracks)
#queryTrackList = [Track(x.trackName, x.title) for x in queryGSuite.allTracks()]
elif genomicRegions == 'Hyperbrowser repository':
selectedGenRegTrack = TfbsTrackNameMappings.getTfbsTrackNameMappings(
genome)[genomicRegionsTracks]
if isinstance(selectedGenRegTrack, dict):
genElementTrackName = selectedGenRegTrack.values()
else:
genElementTrackName = selectedGenRegTrack
elif genomicRegions == 'Hyperbrowser repository (cell-type-specific)':
genElementTrackName = ['Private', 'Antonio'
] + genomicRegionsTracks.split(':')
else:
return
# Get TF track names:
if isinstance(tfTracks, dict):
selectedTfTracks = [
key for key, val in tfTracks.iteritems() if val == 'True'
]
else:
selectedTfTracks = [tfTracks]
queryTrackTitle = '--'.join(genElementTrackName)
trackTitles = [queryTrackTitle]
tracks = [Track(genElementTrackName, trackTitle=queryTrackTitle)]
for i in selectedTfTracks:
if sourceTfs == 'Hyperbrowser repository':
tfTrackName = TfTrackNameMappings.getTfTrackNameMappings(
genome)[sourceTfsDetails] + [i]
tracks.append(
Track(tfTrackName,
trackTitle=tfTrackName[len(tfTrackName) - 1]))
trackTitles.append(tfTrackName[len(tfTrackName) - 1])
else:
tfTrackName = i.split(':')
queryGSuite = getGSuiteFromGalaxyTN(sourceTfsDetails)
for x in queryGSuite.allTracks():
selectedTrackNames = (':'.join(x.trackName))
if i == selectedTrackNames:
tracks.append(Track(x.trackName, x.title))
trackTitles.append(x.trackName[-1])
# queryGSuite = getGSuiteFromGalaxyTN(sourceTfsDetails)
# tfTrackName = [x.trackName for x in queryGSuite.allTracks()] + [i]
# tracks += [Track(x.trackName, x.title) for x in queryGSuite.allTracks()]
# trackTitles += tfTrackName
# print tfTrackName
# print tracks
# print trackTitles
trackTitlesForStat = trackTitles
trackTitles = CommonConstants.TRACK_TITLES_SEPARATOR.join(trackTitles)
##first statistic for Q2
resultsForStatistics = OrderedDict()
similarityFunc = [ #GSuiteStatUtils.T7_RATIO_OF_OBSERVED_TO_EXPECTED_OVERLAP,
GSuiteStatUtils.T5_RATIO_OF_OBSERVED_TO_EXPECTED_OVERLAP
]
for similarityStatClassName in similarityFunc:
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(regSpec,
binSpec,
genome=genome)
mcfdrDepth = AnalysisDefHandler(
REPLACE_TEMPLATES['$MCFDR$']).getOptionsAsText().values()[0][0]
analysisDefString = REPLACE_TEMPLATES[
'$MCFDR$'] + ' -> GSuiteSimilarityToQueryTrackRankingsAndPValuesWrapperStat'
analysisSpec = AnalysisDefHandler(analysisDefString)
analysisSpec.setChoice('MCFDR sampling depth', mcfdrDepth)
analysisSpec.addParameter('assumptions',
'PermutedSegsAndIntersegsTrack_')
analysisSpec.addParameter(
'rawStatistic', GSuiteStatUtils.
PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similarityStatClassName])
analysisSpec.addParameter(
'pairwiseStatistic', GSuiteStatUtils.
PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similarityStatClassName]
) #needed for call of non randomized stat for assertion
analysisSpec.addParameter('tail', 'more')
analysisSpec.addParameter('trackTitles',
trackTitles) #that need to be string
analysisSpec.addParameter('queryTracksNum', str(len(tracks)))
results = doAnalysis(analysisSpec, analysisBins,
tracks).getGlobalResult()
if not similarityStatClassName in resultsForStatistics:
resultsForStatistics[similarityStatClassName] = {}
resultsForStatistics[similarityStatClassName] = results
keyTitle = [
#'Normalized ratio of observed to expected overlap (normalized Forbes similarity measure)',
'Ratio of observed to expected overlap (Forbes similarity measure)'
]
# 'Normalized Forbes coefficient: ratio of observed to expected overlap normalized in relation to the reference GSuite',
# 'Forbes coefficient: ratio of observed to expected overlap'
keyTitle = [
#GSuiteStatUtils.T7_RATIO_OF_OBSERVED_TO_EXPECTED_OVERLAP,
GSuiteStatUtils.T5_RATIO_OF_OBSERVED_TO_EXPECTED_OVERLAP
]
resultDict = AllTfsOfRegions.countStatistics(similarityFunc, choices,
genome, tracks,
trackTitlesForStat)
resultDictShow = AllTfsOfRegions.countStatisticResults(
resultDict, keyTitle, trackTitlesForStat)
# print resultsForStatistics
'''selectedTrackNames = []
if sourceTfs == 'History (user-defined)':
if selectedTfTracks.split(":")[1] == "gsuite":
gSuite = getGSuiteFromGalaxyTN(selectedTfTracks)
for track in gSuite.allTracks():
selectedTrackNames.append(track.trackName)
else:
galaxyTN = selectedTfTracks.split(':')
gRegTrackName = ExternalTrackManager.getPreProcessedTrackFromGalaxyTN(genome, galaxyTN)
selectedTrackNames.append(gRegTrackName)
else:'''
tfNameList = []
#Intersection between TF Tracks and selected region (Table 1):
n = 0
allTargetBins = []
alltfNames = []
table1 = []
for i in selectedTfTracks:
n = n + 1
#newGalaxyFn = galaxyFn.split(".")[0] + str(n) + "." + "dat"
if sourceTfs == 'Hyperbrowser repository':
tfTrackName = TfTrackNameMappings.getTfTrackNameMappings(
genome)[sourceTfsDetails] + [i]
else:
tfTrackName = i.split(':')
tfTrackName.pop(0)
#tfIntersection.expandReferenceTrack(upFlankSize, downFlankSize)
tfIntersection = TrackIntersection(genome, genElementTrackName,
tfTrackName, galaxyFn, str(n))
regFileNamer = tfIntersection.getIntersectedRegionsStaticFileWithContent(
)
targetBins = tfIntersection.getIntersectedReferenceBins()
#regSpec, targetBins = UserBinSelector.getRegsAndBinsSpec(choices)
tfHits = [i] * len(targetBins)
fixedTargetBins = [str(a).split(" ")[0] for a in targetBins]
extendedTargetBins = [
list(a) for a in zip(fixedTargetBins, tfHits)
]
allTargetBins = allTargetBins + extendedTargetBins
tfName = i
alltfNames = alltfNames + [tfName]
# Save output table:
tfNameList.append(tfName)
line = [tfName] + [len(targetBins)] + [
regFileNamer.getLink('Download bed-file')
] + [
regFileNamer.getLoadToHistoryLink('Send bed-file to History')
]
table1 = table1 + [line]
# Computing totals:
fullCase = ','.join(alltfNames)
firstColumn = [item[0] for item in allTargetBins]
uniqueAllTargetBins = list(set(firstColumn))
# Group TFs by bound region:
d1 = defaultdict(list)
for k, v in allTargetBins:
d1[k].append(v)
allTFTargetBins = dict((k, ','.join(v)) for k, v in d1.iteritems())
allTFTargetList = []
fullCaseTFTargetList = []
for key, value in allTFTargetBins.iteritems():
allTFTargetList = allTFTargetList + [[key, value]]
if value == fullCase:
fullCaseTFTargetList = fullCaseTFTargetList + [[key, value]]
analysis3 = TrackIntersection.getFileFromTargetBins(
allTFTargetList, galaxyFn, str(3))
analysis4 = TrackIntersection.getFileFromTargetBins(
fullCaseTFTargetList, galaxyFn, str(4))
# Print output to table:
title = 'TF targets and co-occupancy of ' + genElementTrackName[
-1] + ' genomic regions'
htmlCore = HtmlCore()
pf = plotFunction(tableId='resultsTable')
htmlCore.begin()
htmlCore.header(title)
htmlCore.divBegin('resultsDiv')
htmlCore.line(pf.createButton(bText='Show/Hide more results'))
# htmlCore.tableHeader(['Transcription Factor', 'Normalized ratio of observed to expected overlap (normalized Forbes similarity measure) -- Similarity to genomic regions track', 'Normalized ratio of observed to expected overlap (normalized Forbes similarity measure) -- p-value','Ratio of observed to expected overlap (Forbes similarity measure) -- Similarity to genomic regions track', 'Ratio of observed to expected overlap (Forbes similarity measure) -- p-value', 'Number of TF-Target Track Regions', 'File of TF Target Regions', 'File of TF Target Regions', 'Number of TF-co-occupied Regions', 'File of TF co-occupied Regions', 'File of TF co-occupied Regions', 'Rank of TF co-occupancy motifs', 'Rank of TF co-occupancy motifs'], sortable=True, tableId='resultsTable')
#previous ordering
# htmlCore.tableHeader(['Transcription Factor', 'Normalized Forbes index --overlap score',
# 'Normalized Forbes index --p-value',
# 'Forbes index --overlap score', 'Forbes index --p-value',
# 'Number of TF-Target Track Regions', 'File of TF Target Regions',
# 'File of TF Target Regions', 'Number of target track regions occupied by this TF',
# 'File of TF co-occupied Regions', 'File of TF co-occupied Regions',
# 'Rank of TF co-occupancy motifs', 'Rank of TF co-occupancy motifs'],
# sortable=True, tableId='resultsTable')
htmlCore.tableHeader(
[
'Transcription Factor',
'Number of TF-Target Track Regions',
'File of TF Track Regions',
'Number of target track regions occupied by this TF',
'File of TF Target Regions',
'Forbes index --overlap score',
'Forbes index --p-value',
#'Normalized Forbes index --overlap score', 'Normalized Forbes index --p-value',
'File of TF co-occupied Regions',
'Rank of TF co-occupancy motifs'
],
sortable=True,
tableId='resultsTable')
# Adding co-occupancy results to table:
n = 1000
genRegionNumElements = [
int(x) for x in getTrackRelevantInfo.getNumberElements(
genome, genElementTrackName)
]
for key0, it0 in resultsForStatistics.iteritems():
for el in tfNameList:
if el not in it0:
resultsForStatistics[key0][el] = [None, None]
resultsPlotDict = {}
resultPlotCat = []
resultsPlot = []
resultsForStatisticsProper = {}
for key0, it0 in resultsForStatistics.iteritems():
if not key0 in resultsPlotDict:
resultsPlotDict[key0] = {}
resultsPlotPart = []
for key1, it1 in it0.iteritems():
resultsPlotPart.append(it1[0])
if not key1 in resultsForStatisticsProper:
resultsForStatisticsProper[key1] = []
if not key1 in resultsPlotDict[key0]:
resultsPlotDict[key0][key1] = None
for el in it1:
resultsForStatisticsProper[key1].append(el)
resultsPlotDict[key0][key1] = it1[0]
resultPlotCat.append(tfNameList)
resultPlotCat.append(tfNameList)
#resultPlotCatPart = tfNameList
# print resultPlotCatPart
for key0, it0 in resultsPlotDict.iteritems():
resultsPlotPart = []
for el in tfNameList:
if el in it0:
resultsPlotPart.append(it0[el])
else:
resultsPlotPart.append(None)
resultsPlot.append(resultsPlotPart)
for i in table1:
thisCaseTFTargetList = []
for key, value in allTFTargetList:
if i[0] in value and ',' in value:
thisCaseTFTargetList = thisCaseTFTargetList + [[
key, value
]]
n = n + 1
thisAnalysis = TrackIntersection.getFileFromTargetBins(
thisCaseTFTargetList, galaxyFn, str(n))
thisCaseCoCountsList = []
thing = [x[1] for x in thisCaseTFTargetList]
for k in list(set(thing)):
thisCount = thing.count(k)
thisCaseCoCountsList = thisCaseCoCountsList + \
[[k, thisCount, 100*float(thisCount)/float(sum(genRegionNumElements)), 100*float(thisCount)/float(len(thisCaseTFTargetList))]]
thisCaseCoCountsList.sort(key=lambda x: x[2], reverse=True)
n = n + 1
thisCoCountsAnalysis = TrackIntersection.getOccupancySummaryFile(
thisCaseCoCountsList, galaxyFn, str(n))
thisLine = [len(thisCaseTFTargetList)] + \
[thisAnalysis.getLink('Download file')] + [thisAnalysis.getLoadToHistoryLink('Send file to History')] + \
[thisCoCountsAnalysis.getLink('Download file')] + [thisCoCountsAnalysis.getLoadToHistoryLink('Send file to History')]
newLineI = []
tfName = i[0]
newLineI.append(tfName)
for el in resultsForStatisticsProper[tfName]:
newLineI.append(el)
for elN in range(1, len(i)):
newLineI.append(i[elN])
# htmlCore.tableLine(i + thisLine)
# htmlCore.tableHeader(['Transcription Factor', 'Normalized Forbes index --overlap score',
# 'Normalized Forbes index --p-value',
# 'Forbes index --overlap score', 'Forbes index --p-value',
# 'Number of TF-Target Track Regions', 'File of TF Target Regions',
# 'File of TF Target Regions', 'Number of target track regions occupied by this TF',
# 'File of TF co-occupied Regions', 'File of TF co-occupied Regions',
# 'Rank of TF co-occupancy motifs', 'Rank of TF co-occupancy motifs'],
# sortable=True, tableId='resultsTable')
# htmlCore.tableHeader(['Transcription Factor', 'Number of TF-Target Track Regions', 'File of TF Track Regions',
# 'Number of target track regions occupied by this TF', 'File of TF Target Regions',
# 'Forbes index --overlap score', 'Forbes index --p-value',
# 'Normalized Forbes index --overlap score', 'Normalized Forbes index --p-value',
# 'File of TF co-occupied Regions', 'Rank of TF co-occupancy motifs'],
# sortable=True, tableId='resultsTable')
tl = newLineI + thisLine
# previous ordering tl - 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12
# actual ordering - 0, 5, 7, 8, 7, 3, 4, 1, 2, 9, 11
#ordering = [0, 5, 7, 8, 10, 3, 4, 1, 2, 10, 12]
ordering = [0, 3, 5, 6, 8, 1, 2, 8, 10]
#1, 2, => delete
eoList = []
for eo in ordering:
eoList.append(tl[eo])
htmlCore.tableLine(eoList)
totalCoOccupancyTargetList = []
n = 2000
for key, value in allTFTargetList:
n = n + 1
if ',' in value:
totalCoOccupancyTargetList = totalCoOccupancyTargetList + [[
key, value
]]
#newGalaxyFn = galaxyFn.split(".")[0] + str(n) + "." + "dat"
totalCoOccupancyAnalysis = TrackIntersection.getFileFromTargetBins(
totalCoOccupancyTargetList, galaxyFn, str(n))
#line = ['Total reported regions'] + [len(allTargetBins)] + [''] + [''] + [''] + [''] + ['']
#line = ['Full co-occupancy of ' + fullCase] + ['-'] + ['-'] + ['-'] + ['-'] + ['-'] + ['-'] + ['-'] + [len(fullCaseTFTargetList)] + [analysis4.getLink('Download file')] + [analysis4.getLoadToHistoryLink('Send file to History')] + ['-'] + ['-']
line = ['Full co-occupancy of ' + fullCase] + \
['-'] + \
['-'] + \
[len(fullCaseTFTargetList)] + \
['-'] + \
['-'] + \
['-'] + \
[analysis4.getLoadToHistoryLink('Send file to History')] + \
['-']
htmlCore.tableLine(line)
#line = ['Total unique regions'] + ['-'] + ['-'] + ['-'] + ['-'] + [len(allTFTargetList)] + [analysis3.getLink('Download bed-file')] + [analysis3.getLoadToHistoryLink('Send bed-file to History')] + [len(totalCoOccupancyTargetList)] + [totalCoOccupancyAnalysis.getLink('Download file')] + [totalCoOccupancyAnalysis.getLoadToHistoryLink('Send file to History')] + ['-'] + ['-']
line = ['Total unique regions'] + \
[len(allTFTargetList)] + \
['-'] + \
[len(totalCoOccupancyTargetList)] + \
[analysis3.getLoadToHistoryLink('Send bed-file to History')] + \
['-'] +\
['-'] + \
[totalCoOccupancyAnalysis.getLoadToHistoryLink('Send file to History')] + \
['-']
htmlCore.tableLine(line)
htmlCore.tableFooter()
htmlCore.divEnd()
# htmlCore.line(pf.hideColumns(indexList=[2, 4]))
#
sumRes = 0
for r in resultsPlot[0]:
if r != None:
sumRes += r
if sumRes != 0:
vg = visualizationGraphs()
result = vg.drawColumnCharts(
[resultsPlot[0]],
height=300,
categories=resultPlotCat,
legend=False,
addOptions='width: 90%; float:left; margin: 0 4%;',
#titleText=['Overlap between TFs and genomic region using normalized Forbes', 'Overlap between TFs and genomic region using Forbes'],
titleText=[
'Overlap between TFs and genomic region using Forbes'
],
xAxisRotation=90,
xAxisTitle='TF',
yAxisTitle='value')
htmlCore.line(result)
for key0, it0 in resultDictShow.iteritems():
htmlCore.divBegin('resultsDiv' + str(key0))
htmlCore.header(key0)
htmlCore.tableHeader(it0[0],
sortable=True,
tableId='resultsTable' + str(key0))
for elN in range(1, len(it0)):
htmlCore.tableLine(it0[elN])
htmlCore.tableFooter()
htmlCore.divEnd()
htmlCore.hideToggle(styleClass='debug')
htmlCore.end()
print htmlCore
def execute(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.
'''
gSuite = getGSuiteFromGalaxyTN(choices.gsuite)
attributeNames = gSuite.attributes
attributeValCountDict = dict()
for attrName in attributeNames:
attributeValCountDict[attrName] = defaultdict(int)
for gsTrack in gSuite.allTracks():
for attrName in gsTrack.attributes:
attributeValCountDict[attrName][gsTrack.getAttribute(
attrName)] = attributeValCountDict[attrName][
gsTrack.getAttribute(attrName)] + 1
htmlCore = HtmlCore()
htmlCore.begin()
htmlCore.divBegin(divId='results-page')
htmlCore.divBegin(divClass='results-section')
htmlCore.header('Meta-data summary:')
summaryList = []
locationLine = '''Location:<b> %s</b>. %s'''
if gSuite.location == GSuiteConstants.UNKNOWN:
summaryList.append(locationLine % (
gSuite.location,
"The location of the tracks in the GSuite is not specified as remote or local."
))
elif gSuite.location == GSuiteConstants.REMOTE:
summaryList.append(
locationLine %
(gSuite.location,
"The tracks in the GSuite are located in a remote server."))
elif gSuite.location == GSuiteConstants.LOCAL:
summaryList.append(
locationLine %
(gSuite.location,
"The tracks in the GSuite are located on your server."))
elif gSuite.location == GSuiteConstants.MULTIPLE:
summaryList.append(locationLine % (
gSuite.location,
"The tracks in the GSuite are located both on your local server and at a remote location."
))
fileFormatLine = '''File format:<b> %s</b>. %s'''
if gSuite.fileFormat == GSuiteConstants.UNKNOWN:
summaryList.append(fileFormatLine % (
gSuite.fileFormat,
"The file format of the tracks in the GSuite is not specified."
))
elif gSuite.fileFormat == GSuiteConstants.PREPROCESSED:
summaryList.append(fileFormatLine % (
gSuite.fileFormat,
"The tracks in the GSuite are preprocessed and ready for analysis."
))
elif gSuite.fileFormat == GSuiteConstants.PRIMARY:
summaryList.append(fileFormatLine % (
gSuite.fileFormat,
"The tracks in the GSuite can be manipulated, but must be preprocessed first for analysis."
))
elif gSuite.fileFormat == GSuiteConstants.MULTIPLE:
summaryList.append(fileFormatLine % (
gSuite.fileFormat,
"The tracks in the GSuite are both in preprocessed and primary formats. Only preprocessed tracks can be analyzed by HyperBrowser tools."
))
trackTypeLine = '''Track type:<b> %s</b>. %s'''
if gSuite.trackType == GSuiteConstants.UNKNOWN:
summaryList.append(
trackTypeLine %
(gSuite.trackType,
"The track type of the tracks in the GSuite is not specified."
))
elif gSuite.trackType == GSuiteConstants.MULTIPLE:
summaryList.append(
trackTypeLine %
(gSuite.trackType,
"The tracks in the GSuite are of different track types."))
else:
summaryList.append(trackTypeLine % (
gSuite.trackType,
"The tracks in the GSuite are all (subtypes) of the same type."
))
genomeLine = '''Genome:<b> %s</b>. %s'''
if gSuite.genome == GSuiteConstants.UNKNOWN:
summaryList.append(
genomeLine %
(gSuite.genome,
"The genome of the tracks in the GSuite is not specified."))
elif gSuite.fileFormat == GSuiteConstants.MULTIPLE:
summaryList.append(
genomeLine %
(gSuite.genome,
"The tracks in the GSuite are of different genomes."))
else:
summaryList.append(
genomeLine %
(gSuite.genome,
"The tracks in the GSuite come from the same genome."))
htmlCore.unorderedList(summaryList)
if len(attributeNames) > 0:
paragraph2 = '''
There are<b> %s </b>attributes in the GSuite.
For each of the attributes the most frequent value is given in the table below.
''' % str(len(attributeNames))
htmlCore.paragraph(paragraph2)
tableDataDict = OrderedDict()
for attrName in attributeNames:
maxVal, maxCount = max(
attributeValCountDict[attrName].iteritems(),
key=operator.itemgetter(1))
if maxCount == 1:
mostFreqVal = '[All values are unique]'
nrOfOccurrences = 1
elif maxCount == gSuite.numTracks():
mostFreqVal = maxVal
nrOfOccurrences = str(gSuite.numTracks()) + ' [all tracks]'
else:
mostFreqValList = [
x for x, y in
attributeValCountDict[attrName].iteritems()
if y == maxCount
]
mostFreqVal = ' | '.join(mostFreqValList)
nrOfOccurrences = str(maxCount) + ' [tie]' if len(
mostFreqValList) > 1 else maxCount
tableDataDict[attrName] = [mostFreqVal, nrOfOccurrences]
htmlCore.tableFromDictionary(tableDataDict, [
'Attribute name', 'Most frequent value',
'Number of occurrences'
],
sortable=False,
expandable=False)
htmlCore.divEnd()
htmlCore.divEnd()
htmlCore.end()
print htmlCore
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.
'''
cls._setDebugModeIfSelected(choices)
genome = choices.genome
queryGSuite = getGSuiteFromGalaxyTN(choices.queryGSuite)
refGSuite = getGSuiteFromGalaxyTN(choices.refGSuite)
if choices.similarityFunc:
similarityStatClassNameKey = choices.similarityFunc
else:
similarityStatClassNameKey = GSuiteStatUtils.T5_RATIO_OF_OBSERVED_TO_EXPECTED_OVERLAP
isPointsVsSegments, pointsGSuite, segGSuite = cls.isPointsVsSegmentsAnalysis(queryGSuite, refGSuite)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(regSpec, binSpec, genome=genome)
queryTrackList = [Track(x.trackName, x.title) for x in queryGSuite.allTracks()]
refTrackList = [Track(x.trackName, x.title) for x in refGSuite.allTracks()]
queryTrackTitles = CommonConstants.TRACK_TITLES_SEPARATOR.join(
[quote(x.title, safe='') for x in queryGSuite.allTracks()])
refTrackTitles = CommonConstants.TRACK_TITLES_SEPARATOR.join(
[quote(x.title, safe='') for x in refGSuite.allTracks()])
analysisSpec = AnalysisSpec(GSuiteVsGSuiteWrapperStat)
analysisSpec.addParameter('queryTracksNum', str(len(queryTrackList)))
analysisSpec.addParameter('refTracksNum', str(len(refTrackList)))
analysisSpec.addParameter('queryTrackTitleList', queryTrackTitles)
analysisSpec.addParameter('refTrackTitleList', refTrackTitles)
analysisSpec.addParameter('similarityStatClassName',
GSuiteStatUtils.PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similarityStatClassNameKey])
if choices.removeZeroRow:
analysisSpec.addParameter('removeZeroRow', choices.removeZeroRow)
if choices.removeZeroCol:
analysisSpec.addParameter('removeZeroColumn', choices.removeZeroCol)
resultsObj = doAnalysis(analysisSpec, analysisBins, queryTrackList + refTrackList)
results = resultsObj.getGlobalResult()
# baseDir = GalaxyRunSpecificFile([RAW_OVERLAP_TABLE_RESULT_KEY], galaxyFn).getDiskPath()
# rawOverlapHeatmapPresenter = HeatmapFromDictOfDictsPresenter(resultsObj, baseDir,
# 'Overlapping base-pair of tracks from the two suites',
# printDimensions=False)
rawOverlapTableData = results[RAW_OVERLAP_TABLE_RESULT_KEY]
maxRawOverlap, maxROt1, maxROt2 = rawOverlapTableData.getMaxElement()
similarityScoreTableData = results[SIMILARITY_SCORE_TABLE_RESULT_KEY]
maxSimScore, maxSSt1, maxSSt2 = similarityScoreTableData.getMaxElement()
baseDir = GalaxyRunSpecificFile([], galaxyFn=galaxyFn).getDiskPath()
heatmapPresenter = HeatmapFromTableDataPresenter(resultsObj, baseDir=baseDir,
header='Overlapping base-pairs between the tracks of the two suites',
printDimensions=False)
tablePresenter = MatrixGlobalValueFromTableDataPresenter(resultsObj, baseDir=baseDir,
header='Table of overlapping base-pairs between the tracks of the two suites')
core = HtmlCore()
core.begin()
core.divBegin(divId='results-page')
core.divBegin(divId='svs-res-main-div', divClass='svs-res-main')
core.divBegin(divId='raw-overlap-div', divClass='results-section')
core.divBegin(divId='raw-overlap-table', divClass='svs-table-div')
core.header('Base-pair overlaps between the tracks of the two GSuites')
core.paragraph("""From the tracks in the two GSuites the highest base-pair overlap <b>(%s bps)</b>
is observed for the pair of <b>'%s'</b> and <b>'%s'</b>.""" % (maxRawOverlap, maxROt1, maxROt2))
core.divBegin(divId='raw-table-result', divClass='result-div')
core.divBegin(divId='raw-table-result', divClass='result-div-left')
core.line('''Follow the links to view the results in an HTML table
or raw tabular form:''')
core.divEnd()
core.divBegin(divId='raw-table-result', divClass='result-div-right')
core.line(tablePresenter.getReference(RAW_OVERLAP_TABLE_RESULT_KEY))
core.divEnd()#rawoverlap table
core.divEnd()
core.divEnd()
core.divBegin(divId='raw-overlap-heatmap', divClass='svs-heatmap-div')
try:
core.header('Heatmap of base-pair overlaps')
core.divBegin(divId='raw-table-result', divClass='result-div-heatmap')
core.divBegin(divId='raw-table-result', divClass='result-div-left')
core.line('''Follow the links to view the heatmap in the desired format:''')
core.divEnd()
core.divBegin(divId='raw-table-result', divClass='result-div-right')
core.line(heatmapPresenter.getReference(RAW_OVERLAP_TABLE_RESULT_KEY))
core.divEnd()
core.divEnd()
except:
core.line('Heatmap for the base-pair overlaps could not be created.')
core.divEnd()
core.divEnd()
core.divEnd()#rawoverlap heatmap
core.divEnd()#rawoverlap
core.divBegin(divId='sim-score-div', divClass='results-section')
core.divBegin(divId='sim-score-table', divClass='svs-table-div')
core.header('Similarity score between the tracks of the two GSuites measured by %s' % choices.similarityFunc)
core.paragraph("""From the tracks in the two GSuites the highest similarity score <b>(%s)</b>
is observed for the pair of <b>'%s'</b> and <b>'%s'</b>.""" % (maxSimScore, maxSSt1, maxSSt2))
core.divBegin(divId='raw-table-result', divClass='result-div')
core.divBegin(divId='raw-table-result', divClass='result-div-left')
core.line("""Follow the links to view the results in an HTML table or raw tabular form:""")
core.divEnd()
core.divBegin(divId='raw-table-result', divClass='result-div-right')
core.line(tablePresenter.getReference(SIMILARITY_SCORE_TABLE_RESULT_KEY))
core.divEnd()
core.divEnd()
core.divEnd()#simscore table
core.divBegin(divId='sim-score-heatmap', divClass='svs-heatmap-div')
try:
core.header('Heatmap of similarity scores')
core.divBegin(divId='raw-table-result', divClass='result-div-heatmap')
core.divBegin(divId='raw-table-result', divClass='result-div-left')
core.line('''Follow the links to view the heatmap in the desired format:''')
core.divEnd()
core.divBegin(divId='raw-table-result', divClass='result-div-right')
core.line(heatmapPresenter.getReference(SIMILARITY_SCORE_TABLE_RESULT_KEY))
core.divEnd()
core.divEnd()
except:
core.line('Heatmap for the similarity score could not be created.')
core.divEnd()
core.divEnd()
core.divEnd()#simscore heatmap
core.divEnd()#simscore
core.divEnd()#results
# core.paragraph(
# '''Table displaying the number of base-pairs overlapping between the tracks in the two suites:''')
# core.tableFromDictOfDicts(rawOverlapTableData, firstColName='Track title')
# # core.paragraph(rawOverlapHeatmapPresenter.getReference(resDictKey=RAW_OVERLAP_TABLE_RESULT_KEY))
# core.paragraph(
# '''Table displaying the similarity score for the tracks in the two suites as measured by %s:''' % similarityStatClassNameKey)
# core.tableFromDictOfDicts(similarityScoreTableData, firstColName='Track title')
#
core.divEnd()
core.end()
print str(core)
def execute(cls, choices, galaxyFn=None, username=''):
cls._setDebugModeIfSelected(choices)
targetGSuite = getGSuiteFromGalaxyTN(choices.gSuiteFirst)
refGSuite = getGSuiteFromGalaxyTN(choices.gSuiteSecond)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
analysisDef = 'dummy -> RawOverlapStat'
# analysisDef = 'dummy [withOverlaps=yes] -> RawOverlapAllowSingleTrackOverlapsStat'
results = OrderedDict()
for targetTrack in targetGSuite.allTracks():
targetTrackName = targetTrack.title
for refTrack in refGSuite.allTracks():
refTrackName = refTrack.title
if targetTrack.trackName == refTrack.trackName:
# print targetTrack.title
# print targetTrack.trackName
result = DetermineSuiteTracksCoincidingWithAnotherSuite.handleSameTrack(
targetTrack.trackName, regSpec, binSpec,
targetGSuite.genome, galaxyFn)
else:
result = GalaxyInterface.runManual(
[targetTrack.trackName, refTrack.trackName],
analysisDef,
regSpec,
binSpec,
targetGSuite.genome,
galaxyFn,
printRunDescription=False,
printResults=False,
printProgress=False).getGlobalResult()
if targetTrackName not in results:
results[targetTrackName] = OrderedDict()
results[targetTrackName][refTrackName] = result
stat = STAT_OVERLAP_COUNT_BPS
statIndex = STAT_LIST_INDEX[stat]
title = ''
processedResults = []
headerColumn = []
for targetTrackName in targetGSuite.allTrackTitles():
resultRowDict = processRawResults(results[targetTrackName])
resultColumn = []
headerColumn = []
for refTrackName, statList in resultRowDict.iteritems():
resultColumn.append(statList[statIndex])
headerColumn.append(refTrackName)
processedResults.append(resultColumn)
outputTable = {}
for elN in range(0, len(headerColumn)):
outputTable[elN] = {}
outputTable[elN]['id'] = headerColumn[elN]
transposedProcessedResults = [list(x) for x in zip(*processedResults)]
# second question sumSecondgSuite
# first question numSecondgSuite
# fifth question numSecondgSuitePercentage
for i in range(0, len(transposedProcessedResults)):
outputTable[i]['sumSecondgSuite'] = sum(
transposedProcessedResults[i])
if not 'numSecondgSuite' in outputTable[i]:
outputTable[i]['numSecondgSuite'] = 0
for j in range(0, len(transposedProcessedResults[i])):
if transposedProcessedResults[i][j] >= 1:
outputTable[i]['numSecondgSuite'] += 1
else:
outputTable[i]['numSecondgSuite'] += 0
outputTable[i]['numSecondgSuitePercentage'] = float(
outputTable[i]['numSecondgSuite']) / float(
targetGSuite.numTracks()) * 100
from gold.statistic.CountSegmentStat import CountSegmentStat
from gold.statistic.CountPointStat import CountPointStat
from gold.description.TrackInfo import TrackInfo
from gold.statistic.CountStat import CountStat
# third question numPairBpSecondgSuite
# fourth question numFreqBpSecondgSuite
i = 0
for refTrack in refGSuite.allTracks():
formatName = TrackInfo(refTrack.genome,
refTrack.trackName).trackFormatName
analysisDef = CountStat
analysisBins = GalaxyInterface._getUserBinSource(
regSpec, binSpec, refTrack.genome)
results = doAnalysis(AnalysisSpec(analysisDef), analysisBins,
[PlainTrack(refTrack.trackName)])
resultDict = results.getGlobalResult()
if len(resultDict) == 0:
outputTable[i]['numPairBpSecondgSuite'] = None
outputTable[i]['numFreqBpSecondgSuite'] = None
outputTable[i]['numFreqUniqueBpSecondgSuite'] = None
else:
outputTable[i]['numPairBpSecondgSuite'] = resultDict['Result']
if outputTable[i]['numPairBpSecondgSuite'] != 0:
outputTable[i]['numFreqBpSecondgSuite'] = float(
outputTable[i]['sumSecondgSuite']) / float(
outputTable[i]['numPairBpSecondgSuite'])
else:
outputTable[i]['numFreqBpSecondgSuite'] = None
if outputTable[i]['sumSecondgSuite'] != 0:
outputTable[i]['numFreqUniqueBpSecondgSuite'] = float(
outputTable[i]['numPairBpSecondgSuite']) / float(
outputTable[i]['sumSecondgSuite'])
else:
outputTable[i]['numFreqUniqueBpSecondgSuite'] = None
i += 1
# sortTable
outputTableLine = []
for key, item in outputTable.iteritems():
line = [
item['id'], item['numSecondgSuite'], item['sumSecondgSuite'],
item['numPairBpSecondgSuite'], item['numFreqBpSecondgSuite'],
item['numFreqUniqueBpSecondgSuite'],
item['numSecondgSuitePercentage']
]
outputTableLine.append(line)
import operator
outputTableLineSort = sorted(outputTableLine,
key=operator.itemgetter(1),
reverse=True)
tableHeader = [
'Region ID ', 'Number of cases with at least one event ',
'Total number of events', 'Genome coverage (unique bp)',
'Number of events per unique bp', 'Number of unique bp per event',
'Percentage of cases with at least one event'
]
htmlCore = HtmlCore()
htmlCore.begin()
htmlCore.line(
"<b>Identification of genomic elements with high event recurrence</b> "
)
htmlCore.header(title)
htmlCore.divBegin('resultsDiv')
htmlCore.tableHeader(tableHeader,
sortable=True,
tableId='resultsTable')
for line in outputTableLineSort:
htmlCore.tableLine(line)
plotRes = []
plotXAxis = []
for lineInx in range(1, len(outputTableLineSort[0])):
plotResPart = []
plotXAxisPart = []
for lineInxO in range(0, len(outputTableLineSort)):
# if outputTableLineSort[lineInxO][lineInx]!=0 and
# if outputTableLineSort[lineInxO][lineInx]!=None:
plotResPart.append(outputTableLineSort[lineInxO][lineInx])
plotXAxisPart.append(outputTableLineSort[lineInxO][0])
plotRes.append(plotResPart)
plotXAxis.append(plotXAxisPart)
htmlCore.tableFooter()
htmlCore.divEnd()
htmlCore.divBegin('plot', style='padding-top:20px;margin-top:20px;')
vg = visualizationGraphs()
res = vg.drawColumnCharts(
plotRes,
titleText=tableHeader[1:],
categories=plotXAxis,
height=500,
xAxisRotation=270,
xAxisTitle='Ragion ID',
yAxisTitle='Number of cases with at least one event',
marginTop=30,
addTable=True,
sortableAccordingToTable=True,
legend=False)
htmlCore.line(res)
htmlCore.divEnd()
htmlCore.hideToggle(styleClass='debug')
htmlCore.end()
print htmlCore
def getRunDescription(genome, trackNames, analysisDef, ubSource,
revEngBatchLine, urlForTrackAutoSelection, **kwArgs):
# genome = ubSource.genome
assert len(trackNames) == 3
core = HtmlCore()
analysis = Analysis(analysisDef, genome, trackNames[0], trackNames[1],
**kwArgs)
core.header('GENOME')
core.append(GenomeInfo(genome).mainInfo(printEmpty=False))
core.divider()
formatChoices = analysis.getFormatConverterChoicesAsText().items()
tr1FormatChoice, tr2FormatChoice = formatChoices if len(
formatChoices) == 2 else (None, None)
first = True
for tn,label,formatChoice in zip(trackNames, \
['TRACK 1', 'TRACK 2', 'INTENSITY TRACK'], \
[tr1FormatChoice, tr2FormatChoice, None]):
if tn in [None, []]:
continue
if not first:
core.divider()
core.header(label)
trackInfo = TrackInfo(genome, tn)
trackText = ''
if ExternalTrackManager.isHistoryTrack(tn):
assert len(
tn) >= 4, 'Length of external track name < 4: %s' % str(tn)
core.descriptionLine(
'Name',
ExternalTrackManager.extractNameFromHistoryTN(tn) +
' (from history)' + os.linesep)
else:
core.descriptionLine('Name', ':'.join(tn) + os.linesep)
core.append(trackInfo.mainInfo(printEmpty=False))
if formatChoice is not None:
core.descriptionLine('Treated as', formatChoice[1])
first = False
core.divider()
core.header('ANALYSIS')
core.paragraph(''.join(str(analysis).split(':')[1:]))
first = True
for label, choice in analysis.getInterfaceChoicesAsText().items():
if first:
core.divider()
core.header('OPTIONS')
core.descriptionLine(label, choice)
first = False
h0 = analysis.getH0()
if h0 is not None:
core.divider()
core.header('NULL HYPOTHESIS')
core.paragraph(h0)
h1 = analysis.getH1()
if h1 is not None:
core.divider()
core.header('ALTERNATIVE HYPOTHESIS')
core.paragraph(h1)
core.divider()
core.header('ANALYSIS REGIONS')
if hasattr(ubSource, 'description'):
core.paragraph(ubSource.description)
core.divider()
core.header('SOLUTION')
statClass = analysis.getStat()
#One alternative is to put getDescription in MagicStatFactory-hierarchy as class-method, and get real class behind partial-object.
#if isinstance(statClass, functools.partial):
#statClass = statClass.func
#core.paragraph( statClass.getDescription() )
#Chosen alternative is to Instantiate an object, which will automatically give object of real class..
#and then use the following two lines, which will get class in Statistic-hierarchy instead of MagicStatFactory-hierarchy ..
try:
reg = ubSource.__iter__().next()
except:
core.paragraph(
'Solution not relevant, as there are no specified analysis regions..'
)
else:
track1, track2 = analysis.getTracks()
if statClass is None:
core.paragraph(
'Solution not available, due to currently invalid analysis'
)
logMessage('Solution not available, with params: ' +
str([trackNames[0], trackNames[1], analysisDef]),
level=logging.WARN)
else:
statObj = statClass(reg, track1, track2)
statDescr = statObj.getDescription()
replPat = '<a href=' + os.sep.join(
[STATIC_REL_PATH, 'notes', 'stats', '']) + r'\1>note</a>'
statDescr = re.sub('<note>(.*)</note>', replPat, statDescr)
core.paragraph(statDescr)
core.divider()
core.header('TIME OF ANALYSIS')
core.paragraph('Analysis initiated at time: ' +
str(datetime.datetime.now()))
if urlForTrackAutoSelection not in [None, '']:
core.divider()
core.header('URL FOR TRACK AUTOSELECTION')
#urlOptions = '&'.join(['track1=' + quote(':'.join(trackName1)), 'track2=' + quote(':'.join(trackName2))])
#core.paragraph(URL_PREFIX + '/hyper?' + urlOptions)
core.styleInfoBegin(styleClass='break-word')
core.paragraph(urlForTrackAutoSelection)
core.styleInfoEnd()
if revEngBatchLine not in [None, '']:
core.divider()
core.header('CORRESPONDING BATCH COMMAND LINE')
#if any(ExternalTrackManager.isRedirectOrExternalTrack(tn) for tn in [trackName1, trackName2]):
#core.paragraph('Batch-run line not available with tracks from history')
#else:
core.styleInfoBegin(styleClass='break-word')
core.paragraph(revEngBatchLine)
core.styleInfoEnd()
core.divider()
core.header('REFERENCES')
core.paragraph(
'The HyperBrowser system is described in:<br>"Sandve et al., <a href="http://genomebiology.com/2010/11/12/R121/">The Genomic HyperBrowser: inferential genomics at the sequence level</a>, Genome Biol. 2010;11(12):R121'
)
from gold.statistic.RandomizationManagerStat import RandomizationManagerStat
if statClass is not None and RandomizationManagerStat.getMcSamplingScheme(
statClass.keywords) == 'MCFDR':
core.paragraph('The p-values of this analysis were computed using the MCFDR scheme for Monte Carlo based p-value computation'+\
', described in:<br>Sandve et al., <a href="http://bioinformatics.oxfordjournals.org/content/early/2011/10/13/bioinformatics.btr568.long">Sequential Monte Carlo multiple testing</a>, Bioinformatics 2011')
# description = \
#'''
#Run descriptions will be introduced in the next version of HB. <br>
#Below is an example run description, which is a static text unconnected to your choices. The purpose is to get feedback from you on what this should look like:<br>
#Track1 (refseg:genes): Unmarked points (converted from unmarked segments, taking midpoints)<br>
#Track2 (DNA melting:meltmap): Function<br>
#Bins: Chr1, divided into bins of 10 megabases<br>
#Question: Are track1-points occurring with different frequency inside track2-segment than outside?<br>
#Analysis:<br>
#The main result is a p-value resulting from a statistical test connected to the question.<br>
#The null-hypothesis assumes that the track1-points are randomly distributed according to a poisson-distribution, with the same number of points as in the original data. Track2-segment are assumed fixed as they are in the original data. This can be answered by a binomial test. The alternative hypothesis is then that the count of points inside segments has resulted from a different distribution of points, where the points are then either distributed more or less inside segments versus outside. See the note on this question in the user guide for further info.<br>
#'''
return str(core)
def presentResults(self):
"""
:return: Returns html core object
"""
core = HtmlCore()
core.begin()
core.header("Results")
core.divBegin(divClass='resultsExplanation')
core.paragraph('''
The table summarizes the results for each transcription factor and PWM that was analysed. Click on a row for details.
''')
core.divEnd()
core._str += """
<script type="text/javascript" src="https://code.jquery.com/jquery-2.1.4.min.js"></script>
<script>
jQuery(document).ready(function() {
jQuery(".content").hide();
//toggle the componenet with class msg_body
jQuery(".heading").click(function()
{
jQuery(this).next(".content").slideToggle(1);
});
});
</script>
"""
core._str += "<table class='colored bordered'>"
#columns = ["Transcription factor", ""]
#core.tableFromDictionary(rows, columns)
core._str += """
<tr>
<th class='header'>Transcription factor</th>
<!--<th>Peak data</th>-->
<th class='header'>Mofif</th>
<th class='header'>Number of peaks</th>
<th class='header'>Number of peaks with SNP(s)</th>
<th class='header'>Number of changed bindings</th>
<!--<th class='header'>Binding after mutation</th>-->
</tr>
"""
for tf in transcription_factors:
# First print some summary information for this TF
name = tf.name
core._str += "<tr class='heading' style='cursor: pointer;'>"
core._str += "<td>" + name + "</td>"
#core._str += "<td>" + ''.join(tf.name) + "</td>"
core._str += "<td>" + tf.motif.name + "</td>"
core._str += "<td>%d</td>" % (len(tf.peaks))
core._str += "<td>%d</td>" % (len(
[p for p in tf.peaks if p.hasSnps]))
subtable = ""
subtable += "<tr class='content'><td colspan='7'>"
subtable += "<br><h4 style='margin-left: 20px;'>Peaks that intersect with one or more SNPs</h4>"
subtable += "<table border='1' cellpadding='5' style='margin-left: 20px;'>"
subtable += """
<tr>
<th>Position</th>
<th>Best binding before mutation</th>
<th>Best binding after mutation</th>
</tr>
"""
rows_important = []
rows = []
n_gain_loss = 0
for peak in tf.peaks:
important = False
if len(peak.tf.motif) > len(peak.sequence):
continue # Ignore motifs longer than peak sequence (will only occur on test sets
row = ""
row += "<tr>"
row += "<td>%s %d:%d</td>" % (peak.chr, peak.start, peak.end)
#print "<td>%s</td>" % (''.join(peak.sequence))
p = peak.bestBindingPositionBeforeMutation
#print "Sequence: " + str(peak.sequence[p - peak.start : p - peak.start + len(peak.tf.motif)])
row += "<td>On pos %d with score %.10f<br>%s</td>" % (
p, peak.bestBindingScoreBeforeMutation,
prettySequence(
peak, peak.sequence[p - peak.start:p - peak.start +
len(peak.tf.motif)], p))
# Only present binding after if there was a mutation either within the old binding or within a new binding
"""
if peak.hasSnpBetween(peak.bestBindingPositionAfterMutation, peak.bestBindingPositionAfterMutation + len(peak.tf.motif)) or \
peak.hasSnpBetween(peak.bestBindingPositionAfterMutation, peak.bestBindingPositionAfterMutation + len(peak.tf.motif)):
subtable += "<td>On pos %d with score %.10f<br>%s<br>Binding sequence: %s</td>" % (peak.bestBindingPositionAfterMutation, peak.bestBindingScoreAfterMutation, \
prettySequence(peak, peak.bindingSequenceAfterMutations, peak.bestBindingPositionAfterMutation),\
''.join(peak.bindingSequenceAfterMutations))
"""
if peak.hasSnpBetween(peak.start, peak.end):
if peak.bestBindingScoreAfterMutation != peak.bestBindingScoreBeforeMutation:
n_gain_loss += 1
row += "<td><font color='darkgreen'>"
else:
row += "<td><font>"
row += "On position %d with score %.10f<br>%s<br>Binding sequence: %s</font></td>" % (peak.bestBindingPositionAfterMutation, peak.bestBindingScoreAfterMutation, \
prettySequence(peak, peak.bindingSequenceAfterMutations, peak.bestBindingPositionAfterMutation),\
''.join(peak.bindingSequenceAfterMutations))
important = True
else:
row += "<td><font color='#666666'>No change (no point mutations)</font></td>"
row += "</tr>"
if important:
rows_important.append(row)
else:
rows.append(row)
subtable += ''.join(rows_important)
#subtable += ''.join(rows)
subtable += "</table><br><br>"
subtable += "</td></tr>"
if n_gain_loss > 0:
core._str += "<td><b>%d</b></td>" % n_gain_loss
else:
core._str += "<td>%d</td>" % n_gain_loss
core._str += "</tr>"
core._str += subtable
core._str += "</table>"
return core
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.
'''
cls._setDebugModeIfSelected(choices)
# First compute pvalue by running the statistic through a wrapper stat that computes the max per bin
"""
from quick.statistic.RandomizationManagerV3Stat import RandomizationManagerV3Stat
from quick.statistic.CollectionBinnedHypothesisWrapperStat import CollectionBinnedHypothesisWrapperStat
analysisSpec = AnalysisSpec(CollectionBinnedHypothesisWrapperStat)
analysisSpec.addParameter("rawStatistic", "GenericMaxBinValueStat")
analysisSpec.addParameter('perBinStatistic', 'SummarizedStat')
analysisSpec.addParameter('mcSamplerClass', 'NaiveMCSamplingV2Stat')
analysisSpec.addParameter('pairwiseStatistic', 'ProportionCountStat')
analysisSpec.addParameter('summaryFunc', choices.summaryFunc)
analysisSpec.addParameter('evaluatorFunc','evaluatePvalueAndNullDistribution')
analysisSpec.addParameter('tail', 'right-tail')
analysisSpec.addParameter('assumptions', 'RandomGenomeLocationTrack')
analysisSpec.addParameter('maxSamples', 10)
gsuite = getGSuiteFromGalaxyTN(choices.gsuite)
tracks = [Track(x.trackName) for x in gsuite.allTracks()]
regSpec, binSpec = cls.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(regSpec,
binSpec,
choices.genome)
results = doAnalysis(analysisSpec, analysisBins, tracks)
print "<p>Max stat results:</p>"
print results.getGlobalResult()
"""
# Stat question 4
summaryFunc = choices.summaryFunc if choices.summaryFunc else cls.SUMMARY_FUNC_DEFAULT
statTxt = "Average"
if (summaryFunc == "max"): statTxt = "Maximum"
statDesc = 'number of <b>segments</b> per base'
if choices.analysisName == cls.Q2:
statDesc = 'number of <b>base pairs covered by segments</b>'
core = HtmlCore()
core.begin()
core.header("Enrichment of GSuite tracks across regions")
core.divBegin(divClass='resultsExplanation')
core.paragraph(
'The following is a list of all regions (bins) and the <b>' +
statTxt.lower() + '</b> ' + statDesc +
' across the tracks within each region.')
core.divEnd()
if choices.analysisName == cls.Q3:
# Compute p-value per bin
analysisSpec = AnalysisSpec(GSuiteBinEnrichmentPValWrapperStat)
analysisSpec.addParameter('rawStatistic', 'BinSizeStat')
#analysisSpec.addParameter('pairwiseStatistic', 'ProportionElementCountStat')
#analysisSpec.addParameter('pairwiseStatistic', 'ProportionElementCountStat')
#analysisSpec.addParameter('summaryFunc', summaryFunc)
gsuite = getGSuiteFromGalaxyTN(choices.gsuite)
tracks = [Track(x.trackName) for x in gsuite.allTracks()]
regSpec, binSpec = cls.getRegsAndBinsSpec(choices)
from quick.statistic.GenericRelativeToGlobalStat import GenericRelativeToGlobalStatUnsplittable
#analysisSpec.addParameter("globalSource", GenericRelativeToGlobalStatUnsplittable.getGlobalSource('test', choices.genome, False))
analysisSpec.addParameter("globalSource", 'userbins')
analysisBins = GalaxyInterface._getUserBinSource(
regSpec, binSpec, choices.genome)
results_pval = doAnalysis(analysisSpec, analysisBins, tracks)
#print results_pval
analysisSpec = AnalysisSpec(SummarizedWrapperStat)
analysisSpec.addParameter('rawStatistic', 'SummarizedWrapperStat')
countStat = 'ProportionElementCountStat'
if choices.analysisName == cls.Q2:
countStat = 'ProportionCountStat'
# analysisSpec.addParameter('pairwiseStatistic', 'ProportionCountStat')
analysisSpec.addParameter('pairwiseStatistic', countStat)
analysisSpec.addParameter('summaryFunc', summaryFunc)
gsuite = getGSuiteFromGalaxyTN(choices.gsuite)
tracks = [Track(x.trackName) for x in gsuite.allTracks()]
regSpec, binSpec = cls.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(
regSpec, binSpec, choices.genome)
results = doAnalysis(analysisSpec, analysisBins, tracks)
prettyResults = {}
#print results
for key, val in results.iteritems():
if "Result" in val.keys():
if choices.analysisName == cls.Q3:
prettyResults[key] = (val["Result"],
results_pval[key]["Result"])
else:
prettyResults[key] = (val["Result"])
else:
prettyResults[key] = "No result"
topTrackTitle = results.keys()[0]
"""
core.paragraph('''
Suite data is coinciding the most in bin %s
''' % ('test'))
"""
columnNames = ['Bin', 'Representation within the bin']
if choices.analysisName == cls.Q3:
columnNames.append('p-value')
core.divBegin()
if choices.analysisName == cls.Q1:
shortQuestion = cls.Q1_SHORT
elif choices.analysisName == cls.Q2:
shortQuestion = cls.Q2_SHORT
else: # Q3
shortQuestion = cls.Q3_SHORT
visibleRows = 20
makeTableExpandable = len(prettyResults) > visibleRows
addTableWithTabularAndGsuiteImportButtons(
core,
choices,
galaxyFn,
shortQuestion,
tableDict=prettyResults,
columnNames=columnNames,
sortable=True,
presorted=0,
expandable=makeTableExpandable)
core.divEnd()
core.end()
print str(core)
)['no overlapping elements'] and sortedGeSourceHasOverlappingRegions(
gtrackSource):
raise InvalidFormatError(
"Error: genome elements are overlapping while header variable 'no overlapping elements' is True."
)
core.append('Done')
valid = True
except Exception, e:
core.append(str(e))
valid = False
core.styleInfoEnd()
core.divider()
core.header('Conclusion:')
core.styleInfoBegin(
styleClass='donemessage' if valid else 'errormessage')
core.highlight('The GTrack file has %s syntax' %
('valid' if valid else 'invalid'))
core.styleInfoEnd()
core.end()
print str(core)
@staticmethod
def validateAndReturnErrors(choices):
'''
Should validate the selected input parameters. If the parameters are not valid,
an error text explaining the problem should be returned. The GUI then shows this text
to the user (if not empty) and greys out the execute button (even if the text is empty).
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 warnings
#warnings.simplefilter('error')
cls._setDebugModeIfSelected(choices)
similarityStatClassName = choices.similarityFunc if choices.similarityFunc else GSuiteStatUtils.T5_RATIO_OF_OBSERVED_TO_EXPECTED_OVERLAP
summaryFunc = choices.summaryFunc if choices.summaryFunc else cls.SUMMARY_FUNC_DEFAULT
pairwiseStatName = GSuiteStatUtils.PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similarityStatClassName]
gsuite = getGSuiteFromGalaxyTN(choices.gsuite)
tracks = [Track(x.trackName) for x in gsuite.allTracks()]
statTxt = "Average"
if(summaryFunc == "max"): statTxt = "Maximum"
if choices.analysisName == cls.Q2:
mcfdrDepth = choices.mcfdrDepth if choices.mcfdrDepth else AnalysisDefHandler(REPLACE_TEMPLATES['$MCFDR$']).getOptionsAsText().values()[0][0]
# First compute pvalue by running the statistic through a wrapper stat that computes the max per bin
#from quick.statistic.CollectionBinnedHypothesisWrapperStat import CollectionBinnedHypothesisWrapperStat
#analysisSpec = AnalysisSpec(CollectionBinnedHypothesisWrapperStat)
analysisDefString = REPLACE_TEMPLATES['$MCFDRv3$'] + ' -> CollectionBinnedHypothesisWrapperStat'
analysisSpec = AnalysisDefHandler(analysisDefString)
analysisSpec.setChoice('MCFDR sampling depth', mcfdrDepth)
analysisSpec.addParameter("rawStatistic", "GenericMaxBinValueStat")
# analysisSpec.addParameter('perBinStatistic', 'SummarizedStat')
analysisSpec.addParameter('perBinStatistic', 'MultitrackSummarizedInteractionV2Stat')
# analysisSpec.addParameter('mcSamplerClass', 'NaiveMCSamplingV2Stat')
analysisSpec.addParameter('pairwiseStatistic', 'ObservedVsExpectedStat')
analysisSpec.addParameter('summaryFunc', summaryFunc)
# analysisSpec.addParameter('evaluatorFunc','evaluatePvalueAndNullDistribution')
analysisSpec.addParameter('tail', 'right-tail')
analysisSpec.addParameter('assumptions', 'RandomGenomeLocationTrack')
#analysisSpec.addParameter('maxSamples', 10)
analysisSpec.addParameter('multitrackSummaryFunc', summaryFunc)
regSpec, binSpec = cls.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(regSpec,
binSpec,
choices.genome)
results = doAnalysis(analysisSpec, analysisBins, tracks)
results = results.getGlobalResult()
resultsTxt = "The highest ranking bin based on the " + statTxt.lower() + " of the Forbes similarity measure for pairs of tracks within each bin had a score of <b>%.3f</b> with p-value <b>%.6f</b>" % (results["TSMC_GenericMaxBinValueStat"], results['P-value'])
# Stat question 7
core = HtmlCore()
core.begin()
analysisSpec = AnalysisSpec(MultitrackSummarizedInteractionWrapperStat)
#analysisSpec.addParameter('pairwiseStatistic', 'ObservedVsExpectedStat')
analysisSpec.addParameter('pairwiseStatistic', GSuiteStatUtils.PAIRWISE_STAT_LABEL_TO_CLASS_MAPPING[similarityStatClassName])
analysisSpec.addParameter('summaryFunc', summaryFunc)
analysisSpec.addParameter('multitrackSummaryFunc', summaryFunc)
gsuite = getGSuiteFromGalaxyTN(choices.gsuite)
tracks = [Track(x.trackName) for x in gsuite.allTracks()]
regSpec, binSpec = cls.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(regSpec,
binSpec,
choices.genome)
results = doAnalysis(analysisSpec, analysisBins, tracks)
#print '<br>results: ', results, '<br><br>'
prettyResults = OrderedDict()
for key, val in results.iteritems():
if "Result" in val.keys():
prettyResults[key] = val["Result"]
else:
prettyResults[key] = "No result"
core.header(statTxt + " co-occurence between pairs of tracks within each bin")
if choices.analysisName == cls.Q2:
core.paragraph(resultsTxt)
core.divBegin(divClass='resultsExplanation')
core.paragraph('The following is a list of all bins and the <b>' + statTxt.lower() + '</b> co-occurrence of tracks within each bin.')
core.divEnd()
"""
core.paragraph('''
Suite data is coinciding the most in bin %s
''' % ('test'))
"""
visibleRows = 20
makeTableExpandable = len(prettyResults) > visibleRows
columnNames = ['Bin', 'Co-occurrence within the bin']
if choices.analysisName == cls.Q1:
shortQuestion = cls.Q1_SHORT
else:
shortQuestion = cls.Q2_SHORT
addTableWithTabularAndGsuiteImportButtons(
core, choices, galaxyFn, shortQuestion, tableDict=prettyResults,
columnNames=columnNames, sortable=True, presorted=0,
expandable=makeTableExpandable, visibleRows=visibleRows)
core.divEnd()
core.end()
print str(core)
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.
'''
DebugMixin._setDebugModeIfSelected(choices)
genome = choices.genome
gSuite = getGSuiteFromGalaxyTN(choices.gsuite)
# fullCategory = AnalysisManager.combineMainAndSubCategories(choices.analysisCategory, 'Basic')
fullCategory = AnalysisManager.combineMainAndSubCategories(
'Descriptive statistics', 'Basic')
tracks = list(gSuite.allTracks())
analysisName = choices.analysis
# selectedAnalysis = GSuiteSingleValueAnalysisPerTrackTool \
# ._resolveAnalysisFromName(gSuite.genome, fullCategory, tracks[0].trackName, analysisName)
selectedAnalysis = cls.ANALYSIS_PRETTY_NAME_TO_ANALYSIS_SPEC_MAPPING[
choices.analysis]
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(regSpec,
binSpec,
genome=genome)
# paramName, paramValues = selectedAnalysis.getFirstOptionKeyAndValues()
# if paramName and paramValues:
# if len(paramValues) == 1:
# selectedAnalysis.addParameter(paramName, paramValues[0])
# else:
# selectedAnalysis.addParameter(paramName, choices.paramOne)
tableDict = OrderedDict()
for track in tracks:
tableDict[track.title] = OrderedDict()
result = doAnalysis(selectedAnalysis, analysisBins, [track])
resultDict = result.getGlobalResult()
if 'Result' in resultDict:
track.setAttribute(analysisName.lower(),
str(resultDict['Result']))
tableDict[
track.title][analysisName] = strWithNatLangFormatting(
resultDict['Result'])
else:
for attrName, attrVal in resultDict.iteritems():
attrNameExtended = analysisName + ':' + attrName
track.setAttribute(attrNameExtended.lower(), str(attrVal))
tableDict[track.title][
attrNameExtended] = strWithNatLangFormatting(attrVal)
# assert isinstance(resultDict['Result'], (int, basestring, float)), type(resultDict['Result'])
core = HtmlCore()
core.begin()
core.header('Results: ' + analysisName)
def _produceTable(core, tableDict=None, tableId=None):
return core.tableFromDictOfDicts(tableDict,
firstColName='Track title',
tableId=tableId,
expandable=True,
visibleRows=20,
presorted=0)
tableId = 'results_table'
tableFile = GalaxyRunSpecificFile([tableId, 'table.tsv'], galaxyFn)
tabularHistElementName = 'Raw results: ' + analysisName
gsuiteFile = GalaxyRunSpecificFile(
[tableId, 'input_with_results.gsuite'], galaxyFn)
GSuiteComposer.composeToFile(gSuite, gsuiteFile.getDiskPath())
gsuiteHistElementName = \
getGSuiteHistoryOutputName('result', ', ' + analysisName, choices.gsuite)
core.tableWithImportButtons(
tabularFile=True,
tabularFn=tableFile.getDiskPath(),
tabularHistElementName=tabularHistElementName,
gsuiteFile=True,
gsuiteFn=gsuiteFile.getDiskPath(),
gsuiteHistElementName=gsuiteHistElementName,
produceTableCallbackFunc=_produceTable,
tableDict=tableDict,
tableId=tableId)
core.end()
print core
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)
# DebugUtil.insertBreakPoint(username=username, currentUser='******')
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)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
analysisBins = GalaxyInterface._getUserBinSource(regSpec,
binSpec,
genome=genome)
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()])
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(
gsuite, stats, analysisBins, queryTrack=None)
if analysisQuestion == cls.Q1:
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', 'No')
analysisSpec.addParameter('trackTitles', trackTitles)
analysisSpec.addParameter('queryTracksNum', len(tracks))
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 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:
mcfdrDepth = choices.mcfdrDepth if choices.mcfdrDepth else \
AnalysisDefHandler(REPLACE_TEMPLATES['$MCFDR$']).getOptionsAsText().values()[0][0]
analysisDefString = REPLACE_TEMPLATES[
'$MCFDRv3$'] + ' -> GSuiteRepresentativenessOfTracksRankingsAndPValuesWrapperStat'
analysisSpec = AnalysisDefHandler(analysisDefString)
analysisSpec.setChoice('MCFDR sampling depth', mcfdrDepth)
analysisSpec.addParameter('assumptions',
'PermutedSegsAndIntersegsTrack')
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('tail', 'right-tail')
analysisSpec.addParameter('trackTitles', trackTitles)
results = doAnalysis(analysisSpec, analysisBins,
tracks).getGlobalResult()
core = HtmlCore()
gsPerTrackResultsModel = GSuitePerTrackResultModel(
results, [
'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 = results.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(results[topTrackTitle][1]),
strWithNatLangFormatting(results[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=results,
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()
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)
def _buildHtml(self, done):
htmlCore = HtmlCore()
htmlCore.begin(reloadTime=RELOAD_TIME)
htmlCore.divBegin(divId='progress')
runningTimeStr = self._getRunningTimeStr()
htmlCore.header(runningTimeStr)
remainingTime, unknown = self._estimateRemainingTime()
timeRemainingStr = self._getEstimatedTimeRemainingStr(
remainingTime, unknown)
if unknown:
if remainingTime > 0:
timeRemainingStr += '+'
htmlCore.header(timeRemainingStr)
nameCellColSpan = 4 #colspan for the first cell that displays the process name
for progressObj in self._progressObjList:
htmlCore.tableHeader([], tableClass='progress')
htmlCore.tableRowBegin(rowClass='progressRow')
htmlCore.tableCell(progressObj.name, colSpan=nameCellColSpan)
# for i in range(progressObj.status):
# content = ''
# if i == int(progressObj.elementCount / 2):
# content = "%0.2f" % float(progressObj.status) / progressObj.elementCount * 100
# if i == int(progressObj.elementCount / 2 + 1):
# content = '%'
# htmlCore.tableCell(content, cellClass='progressCellDone')
#
# for i in range(progressObj.status, progressObj.elementCount):
# content = ''
# if i == int(progressObj.elementCount / 2):
# content = "%0.2f" % float(progressObj.status) / progressObj.elementCount * 100
# if i == int(progressObj.elementCount / 2 + 1):
# content = '%'
# htmlCore.tableCell(content, cellClass='progressCell')
for i in range(progressObj.elementCount):
content = ''
if i == int(progressObj.elementCount / 2):
content = "%0.2f" % (float(progressObj.status) /
progressObj.elementCount * 100)
if i == int(progressObj.elementCount / 2 + 1):
content = '%'
cellCls = 'progressCellDone' if i < progressObj.status else 'progressCell'
htmlCore.tableCell(content, cellClass=cellCls)
htmlCore.tableRowEnd()
htmlCore.tableFooter()
estimatedRemainingTime = progressObj.estimateRemainingTime()
unknown = estimatedRemainingTime == UNKNOWN_TIME_REMAINING
progressObjInfo = self._getEstimatedTimeRemainingStr(
estimatedRemainingTime, unknown)
htmlCore.paragraph(progressObjInfo)
htmlCore.divEnd()
htmlCore.end(stopReload=done)
return htmlCore
def execute_batch(cls, choices, galaxyFn=None, username=''):
print GalaxyInterface.getHtmlBeginForRuns(galaxyFn)
html = HtmlCore()
html.header('Batch run results')
refSnps = cls.get_ref_snp(choices)
#print refSnps
batchMal = "$Tool[hb_variant_melting_profiles](" + '|'.join(
["'%s'"] * len(choices)) + ")"
cmdList = []
for rs in refSnps:
#if len(rs[4]) > 1:
# rs = list(rs)
# rs[4] = list(rs[4])[0]
# rs = tuple(rs)
fakeChoices = (choices.genome, 'Single',
'__batch__') + rs + choices[8:]
#print rs
cmdList.append(batchMal % fakeChoices)
#print cmdList
GalaxyInterface.runBatchLines(cmdList,
galaxyFn,
username=username,
printResults=False,
printProgress=True)
#print HtmlCore().styleInfoEnd()
results_tsv = GalaxyRunSpecificFile(['results.tsv'], galaxyFn)
results = results_tsv.getFile()
dir = os.path.dirname(results_tsv.getDiskPath())
for i in range(0, len(cmdList)):
header = True
ri = 0
for resultline in open(os.path.join(dir, str(i), 'results.tsv')):
if header:
header = False
if i == 0:
headertxt = '#run\t' + resultline
results.write(headertxt)
html.tableHeader(headertxt.split('\t'))
else:
results.write(str(i) + '\t' + resultline)
if resultline.count('?') == 0:
link = '<a href="%d/html/chart-%d.html">%d (graph)</a>' % (
i, ri, i)
else:
link = str(i)
html.tableLine([link] + resultline.split('\t'))
ri += 1
results.close()
html.tableFooter()
# XXX: temp fix for HB/stable bug
if URL_PREFIX == '/hb':
print '</div>'
print '<p><b>' + results_tsv.getLink('Download results') + '</b></p>'
print html
print GalaxyInterface.getHtmlEndForRuns()
def execute(cls, choices, galaxyFn=None, username=''):
cls._setDebugModeIfSelected(choices)
genome = choices.genome
genomicRegionsSource = choices.genomicRegionsSource
genomicRegions = choices.genomicRegions
#upFlankSize = int(choices.upFlankSize)
#downFlankSize = int(choices.downFlankSize)
sourceTfs = choices.sourceTfs
tfTracks = choices.tfTracks
# Get TF track name:
if sourceTfs == cls.REGIONS_FROM_HISTORY:
galaxyTN = tfTracks.split(':')
tfTrackName = ExternalTrackManager.getPreProcessedTrackFromGalaxyTN(
genome, galaxyTN)
else:
tfTrackName = TfTrackNameMappings.getTfTrackNameMappings(
genome)[sourceTfs] + [tfTracks]
# Get Genomic Regions track names:
selectedTrackNames = []
if isinstance(genomicRegions, dict):
selectedGenRegions = [
key for key, val in genomicRegions.iteritems() if val == 'True'
]
else:
selectedGenRegions = genomicRegions
if genomicRegionsSource == 'Hyperbrowser repository (single tracks)':
for i in selectedGenRegions:
selectedTrackNames.append(
TfbsTrackNameMappings.getTfbsTrackNameMappings(genome)[i])
elif genomicRegionsSource == 'Hyperbrowser repository (cell-specific multi-tracks)':
for i in selectedGenRegions:
genElementGSuiteName = TfbsGSuiteNameMappings.getTfbsGSuiteNameMappings(
genome)[i]
gSuite = getGSuiteFromGSuiteFile(genElementGSuiteName)
for track in gSuite.allTracks():
selectedTrackNames.append(track.trackName)
elif genomicRegionsSource == 'History (user-defined)':
if genomicRegions.split(":")[1] == "gsuite":
gSuite = getGSuiteFromGalaxyTN(selectedGenRegions)
for track in gSuite.allTracks():
selectedTrackNames.append(track.trackName)
else:
galaxyTN = selectedGenRegions.split(':')
gRegTrackName = ExternalTrackManager.getPreProcessedTrackFromGalaxyTN(
genome, galaxyTN)
selectedTrackNames.append(gRegTrackName)
else:
return
#Intersection:
title = 'Targets of ' + tfTrackName[-1] + ' TF track'
htmlCore = HtmlCore()
htmlCore.begin()
htmlCore.header(title)
htmlCore.divBegin('resultsDiv')
htmlCore.tableHeader([
'Genomic Region', 'Number of Target Regions',
'Download bed file of Target Regions', 'Send bed file to history'
],
sortable=True,
tableId='resultsTable')
n = 0
allTargetBins = []
dataY = []
allRefSetNames = []
#print 'all:', selectedTrackNames, '<p>'
#print 'tf:', tfTrackName, '<p>'
for i in selectedTrackNames:
n = n + 1
#newGalaxyFn = galaxyFn.split(".")[0] + str(n) + "." + "dat"
tfIntersection = TrackIntersection(genome, i, tfTrackName,
galaxyFn, str(n))
#tfIntersection.expandReferenceTrack(upFlankSize, downFlankSize)
regFileNamer = tfIntersection.getIntersectedRegionsStaticFileWithContent(
)
targetBins = tfIntersection.getIntersectedReferenceBins()
'''print 'Target Bins = ', targetBins, '<p>'
if genomicRegionsSource=='Hyperbrowser repository (single tracks)':
print '\"', tfTracks, '\" in \"', ":".join((i[len(i)-2],i[len(i)-1])), '":<p>'
elif genomicRegionsSource=='History (user-defined)':
print '\"', tfTracks, '\" in \"', i[len(i)-1], '":<p>'
else:
listGenRegion = i[0].split(":")
maxIndex = len(listGenRegion)-1
print '\"', tfTracks, '\" in \"', ":".join((listGenRegion[maxIndex-1],listGenRegion[maxIndex])), '":<p>'
print '<p>Number of Targets = ', len(targetBins), 'regions.</p>'
print '<p>', regFileNamer.getLink('Download bed-file'), ' of all regions with 1 or more hits.</p>'
print '<p>', regFileNamer.getLoadToHistoryLink('Download bed-file to History'), ' of all regions with 1 or more hits.</p>'
print '<p>==============================================</p>'
#with open(galaxyFn, 'w') as outFile:
#print>>outFile, 'TargetBins=', targetBins, '<p>'
#print >>outFile, selectedGenRegions, '<p>' '''
# Collect all target bins and data to plot:
allTargetBins = allTargetBins + targetBins
dataY = dataY + [
TrackIntersection.prepareDataForPlot(genome, targetBins)
]
refSetName = i[len(i) - 1]
allRefSetNames = allRefSetNames + [refSetName]
# Print output to table:
line = [refSetName] + [len(targetBins)] + [
regFileNamer.getLink('Download bed-file')
] + [
regFileNamer.getLoadToHistoryLink(
'Download bed-file to History')
]
#print line, '<p>'
htmlCore.tableLine(line)
line = ['Total'] + [len(allTargetBins)] + [''] + ['']
dataY = dataY + [
TrackIntersection.prepareDataForPlot(genome, allTargetBins)
]
allRefSetNames = allRefSetNames + ['Total']
htmlCore.tableLine(line)
htmlCore.tableFooter()
htmlCore.divEnd()
htmlCore.hideToggle(styleClass='debug')
htmlCore.end()
print htmlCore
#print 'ALL Target Bins = ', allTargetBins, '<p>'
#print 'dataY = ', dataY, '<p>'
# Plot:
if genome == 'hg19':
chrNames = [
'chr1', 'chr2', 'chr3', 'chr4', 'chr5', 'chr6', 'chr7', 'chr8',
'chr9', 'chr10', 'chr11', 'chr12', 'chr13', 'chr14', 'chr15',
'chr16', 'chr17', 'chr18', 'chr19', 'chr20', 'chr21', 'chr22',
'chrX', 'chrY'
]
if genome == 'mm9':
chrNames = [
'chr1', 'chr2', 'chr3', 'chr4', 'chr5', 'chr6', 'chr7', 'chr8',
'chr9', 'chr10', 'chr11', 'chr12', 'chr13', 'chr14', 'chr15',
'chr16', 'chr17', 'chr18', 'chr19', 'chrX', 'chrY'
]
titleText = 'Targets per Chromosome'
dataX = [[dataY[i][j] for i in range(len(dataY))]
for j in range(len(dataY[0]))]
seriesType = ['column'] * len(dataX)
categories = allRefSetNames
yAxisTitle = 'Number of Targets'
seriesName = chrNames
shared = False
legend = True
xAxisRotation = 0
#print 'dataX = ', dataX, '<p>'
htmlCore = HtmlCore()
htmlCore.begin()
title = 'Targets of ' + tfTrackName[-1] + ' TF track per chromosome'
htmlCore.header(title)
htmlCore.line('<a href="#" id="linkContainer1">Click to see plot</a>')
htmlCore.divBegin(divId='plotDiv', style=' margin: 0 auto')
htmlCore.line(vp.addJSlibs())
htmlCore.line(vp.useThemePlot())
htmlCore.line(vp.addJSlibsExport())
plot = vp.drawChart(dataX,
type='column',
legend=legend,
height=600,
xAxisRotation=xAxisRotation,
seriesType=seriesType,
seriesName=seriesName,
shared=shared,
titleText=titleText,
overMouseAxisX=True,
categories=categories,
showChartClickOnLink=True)
htmlCore.line(plot)
htmlCore.divEnd()
htmlCore.end()
print htmlCore
def execute(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.
'''
# from gold.application.LogSetup import setupDebugModeAndLogging
#setupDebugModeAndLogging()
# targetTrackNames, targetTrackCollection, targetTrackGenome = getGSuiteDataFromGalaxyTN(choices.gSuiteFirst)
# targetTracksDict = OrderedDict(zip(targetTrackNames, targetTrackCollection))
# refTrackNames, refTrackCollection, refTrackCollectionGenome = getGSuiteDataFromGalaxyTN(choices.gSuiteSecond)
# refTracksDict = OrderedDict(zip(refTrackNames, refTrackCollection))
#
targetGSuite = getGSuiteFromGalaxyTN(choices.gSuiteFirst)
refGSuite = getGSuiteFromGalaxyTN(choices.gSuiteSecond)
regSpec, binSpec = UserBinMixin.getRegsAndBinsSpec(choices)
if choices.intraOverlap == TrackCollectionsAnalysis.MERGE_INTRA_OVERLAPS:
analysisDef = 'dummy -> RawOverlapStat'
else:
analysisDef = 'dummy [withOverlaps=yes] -> RawOverlapAllowSingleTrackOverlapsStat'
results = OrderedDict()
# for targetTrackName, targetTrack in targetTracksDict.iteritems():
# for refTrackName, refTrack in refTracksDict.iteritems():
for targetTrack in targetGSuite.allTracks():
targetTrackName = targetTrack.title
for refTrack in refGSuite.allTracks():
refTrackName = refTrack.title
if targetTrack.trackName == refTrack.trackName:
result = TrackCollectionsAnalysis.handleSameTrack(targetTrack.trackName, regSpec, binSpec,
choices.genome, galaxyFn)
else:
result = GalaxyInterface.runManual([targetTrack.trackName, refTrack.trackName],
analysisDef, regSpec, binSpec,
choices.genome, galaxyFn,
printRunDescription=False,
printResults=False).getGlobalResult()
if targetTrackName not in results :
results[targetTrackName] = OrderedDict()
results[targetTrackName][refTrackName] = result
stat = choices.statistic
statIndex = STAT_LIST_INDEX[stat]
title = 'Screening track collections (' + stat + ')'
processedResults = []
headerColumn = []
for targetTrackName in targetGSuite.allTrackTitles():
resultRowDict = processRawResults(results[targetTrackName])
resultColumn = []
headerColumn = []
for refTrackName, statList in resultRowDict.iteritems():
resultColumn.append(statList[statIndex])
headerColumn.append(refTrackName)
processedResults.append(resultColumn)
transposedProcessedResults = [list(x) for x in zip(*processedResults)]
tableHeader = ['Track names'] + targetGSuite.allTrackTitles()
htmlCore = HtmlCore()
htmlCore.begin()
htmlCore.header(title)
htmlCore.divBegin('resultsDiv')
htmlCore.tableHeader(tableHeader, sortable=True, tableId='resultsTable')
for i, row in enumerate(transposedProcessedResults):
line = [headerColumn[i]] + [strWithStdFormatting(x) for x in row]
htmlCore.tableLine(line)
htmlCore.tableFooter()
htmlCore.divEnd()
# #hicharts can't handle strings that contain ' or " as input for series names
targetTrackNames = [x.replace('\'', '').replace('"','') for x in targetGSuite.allTrackTitles()]
refTrackNames = [x.replace('\'', '').replace('"','') for x in refGSuite.allTrackTitles()]
#
# '''
# addColumnPlotToHtmlCore(htmlCore, targetTrackNames, refTrackNames,
# stat, title + ' plot',
# processedResults, xAxisRotation = -45, height=800)
# '''
# '''
# addPlotToHtmlCore(htmlCore, targetTrackNames, refTrackNames,
# stat, title + ' plot',
# processedResults, xAxisRotation = -45, height=400)
# '''
#
from quick.webtools.restricted.visualization.visualizationGraphs import visualizationGraphs
vg = visualizationGraphs()
result = vg.drawColumnChart(processedResults,
height=600,
yAxisTitle=stat,
categories=refTrackNames,
xAxisRotation=90,
seriesName=targetTrackNames,
shared=False,
titleText=title + ' plot',
overMouseAxisX=True,
overMouseLabelX = ' + this.value.substring(0, 10) +')
htmlCore.line(result)
#htmlCore.line(vg.visualizeResults(result, htmlCore))
htmlCore.hideToggle(styleClass='debug')
htmlCore.end()
print htmlCore