def updateSignificant(resultsFN, simulationAverageFN, outFN):
id_avgNum = {}
f = open(simulationAverageFN, 'r')
for line in f:
ls = line.strip().split('\t')
id_avgNum[int(ls[0])] = float(ls[1])
f = open(resultsFN, 'r')
newLines = []
for line in f:
ls = line.strip().split('\t')
id = int(ls[0])
numTargets = float(len(ls[4].split(',')))
try:
numExpected = id_avgNum[id]
except KeyError:
numExpected = 0
sigFlag = 'SIG'
if numTargets < numExpected:
sigFlag = 'NON'
#Do Calculations here
updateVal = sigFlag
#update newLines
newLines.append(cg.appendToLine(line, updateVal, int(8)))
f.close()
#update file
f = open(outFN, 'w')
f.writelines(newLines)
f.close()
def convertFile(fN, inFormat, outFormat, oFN = None):
#get input order and extract info
iO = returnOrderList(inFormat)
oO = returnOrderList(outFormat)
f = open(fN, 'r')
newLines = []
for line in f:
ls = line.strip().split(iO[5])
chrom, strand, start, end = ls[iO[0]], ls[iO[1]], ls[iO[2]], ls[iO[3]]
#switch to appropriate chromosome type if needed
if len(chrom) == 1:
chrom = oO[6] + chrom
#switch strand if need be
if oO[4] == 0:
if strand == '1' or strand == '+':
strand = '1'
else:
strand = '-1'
else:
if strand == '1' or strand == '+':
strand = '+'
else:
strand = '-'
#construct new Line
newLine = '\n'
newLine = cg.appendToLine(newLine, chrom, oO[0], sep = oO[5])
newLine = cg.appendToLine(newLine, strand, oO[1], sep = oO[5])
newLine = cg.appendToLine(newLine, start, oO[2], sep = oO[5])
newLine = cg.appendToLine(newLine, end, oO[3], sep = oO[5])
newLines.append(newLine)
f.close()
#output file
f = open(fN + '.' + outFormat, 'w')
f.writelines(newLines)
f.close()
def convertFile(fN, inFormat, outFormat, oFN=None):
#get input order and extract info
iO = returnOrderList(inFormat)
oO = returnOrderList(outFormat)
f = open(fN, 'r')
newLines = []
for line in f:
ls = line.strip().split(iO[5])
chrom, strand, start, end = ls[iO[0]], ls[iO[1]], ls[iO[2]], ls[iO[3]]
#switch to appropriate chromosome type if needed
if len(chrom) == 1:
chrom = oO[6] + chrom
#switch strand if need be
if oO[4] == 0:
if strand == '1' or strand == '+':
strand = '1'
else:
strand = '-1'
else:
if strand == '1' or strand == '+':
strand = '+'
else:
strand = '-'
#construct new Line
newLine = '\n'
newLine = cg.appendToLine(newLine, chrom, oO[0], sep=oO[5])
newLine = cg.appendToLine(newLine, strand, oO[1], sep=oO[5])
newLine = cg.appendToLine(newLine, start, oO[2], sep=oO[5])
newLine = cg.appendToLine(newLine, end, oO[3], sep=oO[5])
newLines.append(newLine)
f.close()
#output file
f = open(fN + '.' + outFormat, 'w')
f.writelines(newLines)
f.close()
def updateTargetsExpression(resultsFN, targetsFN, inputPosition,
updatePosition, outFN):
#load target expression dict
f = open(targetsFN, 'r')
targetsDict = {} # tID: eLevel
for line in f:
targetsDict[int(line.strip().split('\t')[0])] = int(
line.strip().split('\t')[2])
f.close()
#For each sRNA, get target Expression.
f = open(resultsFN, 'r')
newLines = []
for line in f:
targets = line.strip().split('\t')[int(inputPosition)]
targets = targets.strip().split(',')
maxExpressionLevel = 0
totalExpressionLevel = 0
for tID in targets:
tID = int(tID)
tExpressionLevel = targetsDict[tID]
totalExpressionLevel += targetsDict[tID]
if tExpressionLevel > maxExpressionLevel:
maxExpressionLevel = tExpressionLevel
#update newLines
newLine = cg.appendToLine(line, maxExpressionLevel,
int(updatePosition))
newLines.append(
cg.appendToLine(newLine, totalExpressionLevel,
int(updatePosition) + 1))
f.close()
#update file
f = open(outFN, 'w')
f.writelines(newLines)
f.close()
def addFiller(fN, filler, zeroPosition, outFN):
filler = str(filler)
f = open(fN, "r")
newLines = []
for line in f:
newLines.append(bioLibCG.appendToLine(line, filler, int(zeroPosition)))
f.close()
fOut = open(outFN, "w")
fOut.writelines(newLines)
fOut.close()
def addFiller(fN, filler, zeroPosition, outFN):
filler = str(filler)
f = open(fN, 'r')
newLines = []
for line in f:
newLines.append(bioLibCG.appendToLine(line, filler, int(zeroPosition)))
f.close()
fOut = open(outFN, 'w')
fOut.writelines(newLines)
fOut.close()
def transcriptSetOverlapDegFileHitmap(degFile, runningChrom, runningStrand):
geneSetFN = '/home/chrisgre/dataSources/known/Human/geneSets/ensemblAllTranscripts.tsv'
allExons = cgGenes.createGeneSetFromFile(geneSetFN)
transcriptTccs = []
for gene in allExons.set.values():
for transcript in gene.transcripts:
transcriptTccs.append(transcript.tcc)
#create hitmap
coordSet = set()
for tcc in transcriptTccs:
chrom, strand, start, end = cg.tccSplit(tcc)
if chrom != runningChrom:
continue
if strand != runningStrand:
continue
for i in range(start, end + 1):
coordSet.add(i)
#find overlapping degTccs
print 'done creating hitmap'
f = open(degFile, 'r')
newLines = []
for line in f:
ls = line.strip().split('\t')
degTcc = cg.convertToAS(ls[1])
chrom, strand, start, end = cg.tccSplit(degTcc)
if chrom != runningChrom:
continue
if strand != runningStrand:
continue
inTran = '0'
for i in xrange(start, end + 1):
if i in coordSet:
inTran = '1'
break
#update newLines
newLine = cg.appendToLine(line, inTran, 3)
newLines.append(newLine)
f.close()
f = open(degFile + '.%s.%s' % (runningChrom, runningStrand), 'w')
f.writelines(newLines)
f.close()
def transcriptSetOverlapDegFileHitmap(degFile, runningChrom, runningStrand):
geneSetFN = '/home/chrisgre/dataSources/known/Human/geneSets/ensemblAllTranscripts.tsv'
allExons = cgGenes.createGeneSetFromFile(geneSetFN)
transcriptTccs = []
for gene in allExons.set.values():
for transcript in gene.transcripts:
transcriptTccs.append(transcript.tcc)
#create hitmap
coordSet = set()
for tcc in transcriptTccs:
chrom, strand, start, end = cg.tccSplit(tcc)
if chrom != runningChrom:
continue
if strand != runningStrand:
continue
for i in range(start, end + 1):
coordSet.add(i)
#find overlapping degTccs
print 'done creating hitmap'
f = open(degFile, 'r')
newLines = []
for line in f:
ls = line.strip().split('\t')
degTcc = cg.convertToAS(ls[1])
chrom, strand, start, end = cg.tccSplit(degTcc)
if chrom != runningChrom:
continue
if strand != runningStrand:
continue
inTran = '0'
for i in xrange(start, end + 1):
if i in coordSet:
inTran = '1'
break
#update newLines
newLine = cg.appendToLine(line, inTran, 3)
newLines.append(newLine)
f.close()
f = open(degFile + '.%s.%s' % (runningChrom, runningStrand), 'w')
f.writelines(newLines)
f.close()
def updateTargetsExpression(resultsFN, targetsFN, inputPosition, updatePosition, outFN):
#load target expression dict
f = open(targetsFN, 'r')
targetsDict = {} # tID: eLevel
for line in f:
targetsDict[int(line.strip().split('\t')[0])] = int(line.strip().split('\t')[2])
f.close()
#For each sRNA, get target Expression.
f = open(resultsFN, 'r')
newLines = []
for line in f:
targets = line.strip().split('\t')[int(inputPosition)]
targets = targets.strip().split(',')
maxExpressionLevel = 0
totalExpressionLevel = 0
for tID in targets:
tID = int(tID)
tExpressionLevel = targetsDict[tID]
totalExpressionLevel += targetsDict[tID]
if tExpressionLevel > maxExpressionLevel:
maxExpressionLevel = tExpressionLevel
#update newLines
newLine = cg.appendToLine(line, maxExpressionLevel, int(updatePosition))
newLines.append(cg.appendToLine(newLine, totalExpressionLevel, int(updatePosition) + 1))
f.close()
#update file
f = open(outFN, 'w')
f.writelines(newLines)
f.close()
def addFiller(fN, filler, zeroPosition, outFN):
filler = str(filler)
idFlag = False
if filler == 'ID':
idFlag = True
f = open(fN, 'r')
newLines = []
i = 0
for line in f:
if idFlag:
filler = str(i)
newLines.append(bioLibCG.appendToLine(line, filler, int(zeroPosition)))
i += 1
f.close()
fOut = open(outFN, 'w')
fOut.writelines(newLines)
fOut.close()
def transcriptSetOverlapDegFile(degFile):
geneSetFN = '/home/chrisgre/dataSources/known/Human/geneSets/ensemblAllTranscripts.tsv'
allExons = cgGenes.createGeneSetFromFile(geneSetFN)
#get degradome TCCS
#note that you need to test the AS peaks, this is the location of the targetted transcript
degTccs = []
f = open(degFile, 'r')
for line in f:
ls = line.strip().split('\t')
degTccs.append(ls[1])
f.close()
degTccs = [cg.convertToAS(x) for x in degTccs]
#find all overlapping exons/transcripts, then all results sequences that overlap exons
overlappingExons = allExons.transcriptOverlaps(degTccs)
#print len(overlappingExons), "num of overlapping exons"
overlappingExonTccs = [x.tcc for x in overlappingExons]
overlappingDegTccs = compare.compareTwoTcc(degTccs, overlappingExonTccs, 1)
f = open(degFile, 'r')
newLines = []
for line in f:
degTcc = cg.convertToAS(ls[1])
inTran = '0'
if degTcc in overlappingDegTccs:
inTran = '1'
#update newLines
newLine = cg.appendToLine(line, inTran, 3)
f.close()
def transcriptSetOverlapDegFile(degFile):
geneSetFN = '/home/chrisgre/dataSources/known/Human/geneSets/ensemblAllTranscripts.tsv'
allExons = cgGenes.createGeneSetFromFile(geneSetFN)
#get degradome TCCS
#note that you need to test the AS peaks, this is the location of the targetted transcript
degTccs = []
f = open(degFile, 'r')
for line in f:
ls = line.strip().split('\t')
degTccs.append(ls[1])
f.close()
degTccs = [cg.convertToAS(x) for x in degTccs]
#find all overlapping exons/transcripts, then all results sequences that overlap exons
overlappingExons = allExons.transcriptOverlaps(degTccs)
#print len(overlappingExons), "num of overlapping exons"
overlappingExonTccs = [x.tcc for x in overlappingExons]
overlappingDegTccs = compare.compareTwoTcc(degTccs, overlappingExonTccs, 1)
f = open(degFile, 'r')
newLines = []
for line in f:
degTcc = cg.convertToAS(ls[1])
inTran = '0'
if degTcc in overlappingDegTccs:
inTran = '1'
#update newLines
newLine = cg.appendToLine(line, inTran, 3)
f.close()
def updateSignificant(resultsFN, simulationAverageFN, outFN):
id_avgNum = {}
f = open(simulationAverageFN, 'r')
for line in f:
ls = line.strip().split('\t')
id_avgNum[int(ls[0])] = float(ls[1])
f = open(resultsFN, 'r')
newLines = []
for line in f:
ls = line.strip().split('\t')
id = int(ls[0])
numTargets = float(len(ls[4].split(',')))
try:
numExpected = id_avgNum[id]
except KeyError:
numExpected = 0
sigFlag = 'SIG'
if numTargets < numExpected:
sigFlag = 'NON'
#Do Calculations here
updateVal = sigFlag
#update newLines
newLines.append(cg.appendToLine(line, updateVal, int(8)))
f.close()
#update file
f = open(outFN, 'w')
f.writelines(newLines)
f.close()
def updateReadDensity(tType, cName):
#go through wig each chromosome and check the mature seqs
mainConf = cgConfig.cgConfig('Main.conf')
conf = cgConfig.getConfig(cName)
organism = conf.conf['organism']
wigFolder = mainConf.conf['wig%s' % organism]
newLines = []
#Differentiate between exon or intron...
if tType == 'E':
pFileName = conf.conf['resultsExons']
elif tType == 'I':
pFileName = conf.conf['resultsIntrons']
else:
print 'READ UPDATE FAIL'
print ' Updating Read Density:', tType
#get read density for each line...
print ' calculating hits for mature seqs'
#calculate total hits per mature
mirFile = open(pFileName, 'r')
for line in mirFile:
mTcc = line.strip().split('\t')[1]
mirID = line.strip().split('\t')[0]
tccStretch = cgPeaks.stretch(mTcc, cName)
highestHit = 0
for i in range(int(mTcc.split(':')[2]), int(mTcc.split(':')[3])):
if i in tccStretch.profile:
if tccStretch.profile[i] > highestHit:
highestHit = tccStretch.profile[i]
newLines.append(cg.appendToLine(line, str(highestHit), 11))
mirFile.close()
print 'Writing New File'
#write new results file
outFile = open(pFileName, 'w')
for line in newLines:
outFile.write(line)
outFile.close()
####NOW UPDATE HIGHEST HIT PER CLUSTER####
clusterCount = {}
pFile = open(pFileName, 'r')
for line in pFile:
predictionCount = int(line.strip().split('\t')[11])
CID = line.strip().split('\t')[7]
if CID in clusterCount:
if clusterCount[CID] < predictionCount:
clusterCount[CID] = predictionCount
else:
clusterCount[CID] = predictionCount
pFile.close()
#update the file --> cluster small count
newLines = []
predFile = open(pFileName, 'r')
for line in predFile:
CID = line.strip().split('\t')[7]
numMax = clusterCount[CID]
newLines.append(cg.appendToLine(line, str(numMax), 12))
predFile.close()
#sort newLines by clusterID
sortDict = {}
CIDs = []
for line in newLines:
CID = int(line.strip().split('\t')[7])
if CID not in CIDs:
CIDs.append(CID)
if CID in sortDict:
sortDict[CID].append(line)
else:
sortDict[CID] = [line]
CIDs.sort()
newLines = []
for CID in CIDs:
for line in sortDict[CID]:
newLines.append(line)
#write new File
newFile = open(pFileName, 'w')
for line in newLines:
newFile.write(line)
newFile.close()
def updateReadDensity(tType):
#go through wig each chromosome and check the mature seqs
mainConf = cgConfig.cgConfig('Main.conf')
conf = cgConfig.cgConfig()
organism = conf.conf['organism']
wigFolder = mainConf.conf['wig%s' % organism]
newLines = []
if tType == 'E':
pFileName = conf.conf['resultsExons']
elif tType == 'I':
pFileName = conf.conf['resultsIntrons']
else:
print 'READ UPDATE FAIL'
print ' Updating Read Density:', tType
for wigFileN in cg.recurseDir(wigFolder, end='.wig'):
#init
chrom = wigFileN.strip().split('.')[-2]
strand = wigFileN.strip().split('.')[-4]
wigFile = open(wigFileN, 'r')
mirFile = open(pFileName, 'r')
print wigFileN
#get rid of header
wigFile.readline()
print ' populating hitmap'
#populate hitmap
wigMap = {}
for line in wigFile:
value = int(line.strip().split('\t')[3].split('.')[0])
if value > 0:
start = int(line.strip().split('\t')[1])
end = int(line.strip().split('\t')[2])
for i in range(start, end):
wigMap[i] = value
wigFile.close()
print ' calculating hits for mature seqs'
#calculate total hits per mature
for line in mirFile:
mTcc = line.strip().split('\t')[1]
mirID = line.strip().split('\t')[0]
if (mTcc.split(':')[0] == chrom) and (mTcc.split(':')[1]
== strand):
#if mirID == '26477.30.106643972': print 'Starting Total Count'
highestHit = 0
for i in range(int(mTcc.split(':')[2]),
int(mTcc.split(':')[3])):
#if mirID == '26477.30.106643972': print ' ', i
if i in wigMap:
if wigMap[i] > highestHit:
highestHit = wigMap[i]
#if mirID == '26477.30.106643972': print ' ', i, totalHits, wigMap[i]
newLines.append(cg.appendToLine(line, str(highestHit), 11))
mirFile.close()
print 'Writing New File'
#write new results file
outFile = open(pFileName, 'w')
for line in newLines:
outFile.write(line)
outFile.close()
####NOW UPDATE HIGHEST HIT PER CLUSTER####
clusterCount = {}
pFile = open(pFileName, 'r')
for line in pFile:
predictionCount = int(line.strip().split('\t')[11])
CID = line.strip().split('\t')[7]
if CID in clusterCount:
if clusterCount[CID] < predictionCount:
clusterCount[CID] = predictionCount
else:
clusterCount[CID] = predictionCount
pFile.close()
#update the file --> cluster small count
newLines = []
predFile = open(pFileName, 'r')
for line in predFile:
CID = line.strip().split('\t')[7]
numMax = clusterCount[CID]
newLines.append(cg.appendToLine(line, str(numMax), 12))
predFile.close()
#sort newLines by clusterID
sortDict = {}
CIDs = []
for line in newLines:
CID = int(line.strip().split('\t')[7])
if CID not in CIDs:
CIDs.append(CID)
if CID in sortDict:
sortDict[CID].append(line)
else:
sortDict[CID] = [line]
CIDs.sort()
newLines = []
for CID in CIDs:
for line in sortDict[CID]:
newLines.append(line)
#write new File
newFile = open(pFileName, 'w')
for line in newLines:
newFile.write(line)
newFile.close()
print ' calculating hits for mature seqs'
#calculate total hits per mature
for line in mirFile:
mTcc = line.strip().split('\t')[1]
mirID = line.strip().split('\t')[0]
if (mTcc.split(':')[0] == chrom) and (mTcc.split(':')[1] == strand):
#if mirID == '26477.30.106643972': print 'Starting Total Count'
totalHits = 0
for i in range(int(mTcc.split(':')[2]), int(mTcc.split(':')[3])):
#if mirID == '26477.30.106643972': print ' ', i
if i in wigMap:
totalHits += wigMap[i]
#if mirID == '26477.30.106643972': print ' ', i, totalHits, wigMap[i]
newLines.append(cg.appendToLine(line, str(totalHits), 11))
mirFile.close()
print 'Writing New File'
#write new results file
outFile = open(pFileName, 'w')
for line in newLines:
outFile.write(line)
outFile.close()
####NOW UPDATE MAX HITS PER CLUSTER####
clusterCount = {}
pFile = open(pFileName, 'r')
print 'Total', numT
print 'A', aPass
print 'B', bPass
print 'C', cPass
#output results to a file for R
outFile = open('mousePeaksResults.R.data', 'w')
outFile.write('tcc\tpeakOne\tpeakTwo\tdistance\tratio\tmax\thRatio\n')
for peak in bestCombos:
outFile.write('%s\t%s\t%s\t%s\t%s\t%s\n' %
(peak[0], peak[1], peak[2], peak[3], peak[4], peak[5]))
outFile.close()
#now update predFile (SLOT 13)
predFile = open(predName, 'r')
newLines = []
for line in predFile:
CID = cg.ss(line)[7]
if peakDict[CID][1] == 'None':
peakInfo = 'None'
else:
peakInfo = '%s:%s:%s:%s' % (
str(peakDict[CID][1][1])[-3:], str(peakDict[CID][1][2])[-3:],
str(peakDict[CID][1][4]).split('.')[0], peakDict[CID][1][5])
newLines.append(cg.appendToLine(line, peakInfo, 13))
predFile.close()
predFile = open(predName, 'w')
predFile.writelines(newLines)
predFile.close()
def updateReadDensity(tType, cName):
#go through wig each chromosome and check the mature seqs
mainConf = cgConfig.cgConfig('Main.conf')
conf = cgConfig.getConfig(cName)
organism = conf.conf['organism']
wigFolder = mainConf.conf['wig%s' % organism]
newLines = []
#Differentiate between exon or intron...
if tType == 'E':
pFileName = conf.conf['resultsExons']
elif tType == 'I':
pFileName = conf.conf['resultsIntrons']
else:
print 'READ UPDATE FAIL'
print ' Updating Read Density:', tType
#get read density for each line...
print ' calculating hits for mature seqs'
#calculate total hits per mature
mirFile = open(pFileName, 'r')
for line in mirFile:
mTcc = line.strip().split('\t')[1]
mirID = line.strip().split('\t')[0]
tccStretch = cgPeaks.stretch(mTcc, cName)
highestHit = 0
for i in range(int(mTcc.split(':')[2]), int(mTcc.split(':')[3])):
if i in tccStretch.profile:
if tccStretch.profile[i] > highestHit:
highestHit = tccStretch.profile[i]
newLines.append(cg.appendToLine(line, str(highestHit), 11))
mirFile.close()
print 'Writing New File'
#write new results file
outFile = open(pFileName, 'w')
for line in newLines:
outFile.write(line)
outFile.close()
####NOW UPDATE HIGHEST HIT PER CLUSTER####
clusterCount = {}
pFile = open(pFileName, 'r')
for line in pFile:
predictionCount = int(line.strip().split('\t')[11])
CID = line.strip().split('\t')[7]
if CID in clusterCount:
if clusterCount[CID] < predictionCount:
clusterCount[CID] = predictionCount
else:
clusterCount[CID] = predictionCount
pFile.close()
#update the file --> cluster small count
newLines = []
predFile = open(pFileName, 'r')
for line in predFile:
CID = line.strip().split('\t')[7]
numMax = clusterCount[CID]
newLines.append(cg.appendToLine(line, str(numMax), 12))
predFile.close()
#sort newLines by clusterID
sortDict = {}
CIDs = []
for line in newLines:
CID = int(line.strip().split('\t')[7])
if CID not in CIDs:
CIDs.append(CID)
if CID in sortDict:
sortDict[CID].append(line)
else:
sortDict[CID] = [line]
CIDs.sort()
newLines = []
for CID in CIDs:
for line in sortDict[CID]:
newLines.append(line)
#write new File
newFile = open(pFileName, 'w')
for line in newLines:
newFile.write(line)
newFile.close()
else:
print 'None'
print timer.split()
#output results to a file for R
outFile = open('mousePeaksResults.R.data', 'w')
outFile.write('tcc\tpeakOne\tpeakTwo\tdistance\tratio\tmax\thRatio\n')
for peak in bestCombos:
outFile.write('%s\t%s\t%s\t%s\t%s\t%s\n' % (peak[0], peak[1],peak[2],peak[3],peak[4],peak[5]))
outFile.close()
#now update predFile (SLOT 13)
predFile = open(predName, 'r')
newLines = []
for line in predFile:
CID = cg.ss(line)[7]
if peakDict[CID][1] == 'None':
peakInfo = 'None'
else:
peakInfo = '%s:%s:%s:%s' % (str(peakDict[CID][1][1])[-3:], str(peakDict[CID][1][2])[-3:], str(peakDict[CID][1][4]).split('.')[0], peakDict[CID][1][5])
newLines.append(cg.appendToLine(line, peakInfo, 13))
predFile.close()
predFile = open(predName, 'w')
predFile.writelines(newLines)
predFile.close()
ensGenes = cgGenes.createGeneSetFromFile(geneSetFolder + '/ensemblAllTranscripts.tsv')
cDesc = {} #CID:gDesc
for CID in cHairs:
tcc = cHairs[CID]
cDesc[CID] = "NONE"
overlappingGenes = ensGenes.geneOverlaps([tcc])
if len(overlappingGenes) > 0:
print overlappingGenes[0].type
cDesc[CID] = overlappingGenes[0].type
f = open(fN, 'r')
newLines = []
for line in f:
CID = line.strip().split('\t')[7]
newLines.append(cg.appendToLine(line, cDesc[CID], 16))
f.close()
f = open(fN + '.FINAL', 'w')
f.writelines(newLines)
f.close()
mConf = c.getConfig('Main.conf')
geneSetFolder = mConf.conf['geneSetsHuman']
fN = '/home/chrisgre/projects/NoncodingHuman/results/NChuman-s3k8b17.results.sorted.introns.sorted'
cHairs = getHairpins.getHairpins(fN)
ensGenes = cgGenes.createGeneSetFromFile(geneSetFolder +
'/ensemblAllTranscripts.tsv')
cDesc = {} #CID:gDesc
for CID in cHairs:
tcc = cHairs[CID]
cDesc[CID] = "NONE"
overlappingGenes = ensGenes.geneOverlaps([tcc])
if len(overlappingGenes) > 0:
print overlappingGenes[0].type
cDesc[CID] = overlappingGenes[0].type
f = open(fN, 'r')
newLines = []
for line in f:
CID = line.strip().split('\t')[7]
newLines.append(cg.appendToLine(line, cDesc[CID], 16))
f.close()
f = open(fN + '.FINAL', 'w')
f.writelines(newLines)
f.close()
def findPeaks(pType, cName=None):
#init
mConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
if pType == 'E':
predName = conf.conf['resultsExonsSorted']
else:
predName = conf.conf['resultsIntronsSorted']
print predName
#make CID:hairpin:peak dictionary
cHairs = getHairpins.getHairpins(predName)
peakDict = {}
for CID in cHairs:
peakDict[CID] = [cHairs[CID], 'None']
timer = cg.cgTimer()
timer.start()
#put peaks in memory
print 'Creating peak data'
peaks = {} # chr:peak:value
for CID in cHairs:
chrom, strand, start, end = cg.tccSplit(cHairs[CID])
tcc = cHairs[CID]
#init dictionary
if chrom not in peaks:
peaks[chrom] = {}
if strand not in peaks[chrom]:
peaks[chrom][strand] = {}
#create peaks for tcc and add to peak dictionary
stretch = cgPeaks.stretch(tcc, cName)
stretch.createPeaks()
for peakCoord in stretch.peaks:
peaks[chrom][strand][peakCoord] = 0
print timer.split()
print 'finding best combos'
bestCombos = []
aPass = 0
bPass = 0
cPass = 0
numT = 0
for CID in peakDict:
cgFlag = False
if CID == '538': cgFlag = True
tcc = peakDict[CID][0]
#print tcc
tccPeaks = []
chrom = cg.ss(tcc, ':')[0]
strand = cg.ss(tcc, ':')[1]
start = int(cg.ss(tcc, ':')[2])
end = int(cg.ss(tcc, ':')[3])
#get all peaks
for i in range(start, end + 1):
if i in peaks[chrom][strand]:
#print ' peak added', i
tccPeaks.append(i)
#Calculate parameters...
pairStrings = [] #used to check if pair already added
peakCombos = []
for x in tccPeaks:
#scan a 30 bp range around this point and find the best roof...
pRange = 30
rTcc = cg.makeTcc(chrom, strand, x, x + 1)
#quickly get max value...kinda a long way to do it but whatever
cProfile = stepVectorScan.profileAroundPoint(rTcc,
1,
cName,
ratio=False)
xval = cProfile[0]
max = xval
highestValueCoord = x
#now make profile for roof...
cProfile = stepVectorScan.profileAroundPoint(rTcc,
pRange,
cName,
ratio=True)
#now get highest stretch length and the rNext coord.
minVal = .80
highest = 0
stretch = 0
startCurrent = None
startFinal = None
endFinal = None
for i in range(1 - pRange, pRange):
if cProfile[i] > minVal:
stretch += 1
if startCurrent == None:
startCurrent = i
else:
if stretch > 0:
if stretch > highest: #stretch ended and was higher than previous
highest = stretch
endFinal = i - 1
startFinal = startCurrent
startCurrent = None
else:
startCurrent = None
stretch = 0
#get +/- 4 value...
val = [1.0, 1.0]
if (startFinal) and (endFinal):
low = startFinal - 4
high = endFinal + 4
if low > (1 - pRange):
if high < pRange:
val[0] = float(cProfile[startFinal - 4])
val[1] = float(cProfile[endFinal + 4])
#fill in other details...
y = 'S'
dist = 'S'
ratio = 'S'
peakCombos.append([tcc, x, y, dist, ratio, max, highest, val])
#print ' ', peakCombos[-1]
#find best combo...
topCombo = None
for combo in peakCombos:
roofLength = combo[6]
dropValue = combo[7][0]
if combo[7][1] > dropValue:
dropValue = combo[7][1]
#print roofLength, dropValue
if 14 < roofLength < 26:
if 0.0 < dropValue < 0.2:
#pick one with rooflength nearest 20:
if topCombo:
if (math.fabs(22 - roofLength)) < (
math.fabs(22 - topCombo[6])):
topCombo = combo
else:
topCombo = combo
if topCombo:
peakDict[CID][1] = topCombo
bestCombos.append(topCombo)
print bestCombos[-1]
else:
#print 'None'
pass
print timer.split()
#now update predFile (SLOT 13)
predFile = open(predName, 'r')
newLines = []
for line in predFile:
CID = cg.ss(line)[7]
if peakDict[CID][1] == 'None':
peakInfo = 'None'
else:
peakInfo = '%s:%s:%s:%s:%s:%s' % (
str(peakDict[CID][1][1])[-3:], 'S', str(
peakDict[CID][1][4]).split('.')[0], peakDict[CID][1][5],
peakDict[CID][1][6], peakDict[CID][1][7])
newLines.append(cg.appendToLine(line, peakInfo, 13))
predFile.close()
predFile = open(predName, 'w')
predFile.writelines(newLines)
predFile.close()
def updateNoise(pType, cName=None):
#init
mainConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
if pType == 'E':
predName = conf.conf['resultsExons']
else:
predName = conf.conf['resultsIntrons']
#populate cid: exon dist
print 'Populating CID/INtron/exon distribution data'
if pType == 'E':
noiseFN = conf.conf['exonNoiseData']
f = open(noiseFN, 'r')
else:
noiseFN = conf.conf['intronNoiseData']
f = open(noiseFN, 'r')
exonDists = {} #cid: [exon dist]
header = f.readline()
order = {} # num:CID
for i, CID in enumerate(header.strip().split('\t')):
order[i] = CID
exonDists[CID] = []
for line in f:
data = line.strip().split('\t')
for i, dataPoint in enumerate(data):
if dataPoint == 'NA' or dataPoint == '':
continue
else:
dataPoint = float(dataPoint)
CID = order[i]
exonDists[CID].append(dataPoint)
#get highest expression level for each cluster
print 'Populating highest expression levels'
predExpression = {} # CID; highest level
exonFile = open(predName, 'r')
for line in exonFile:
CID = line.strip().split('\t')[7]
hDensity = line.strip().split('\t')[12]
predExpression[CID] = hDensity
#get pVals for each CID
print 'Getting pvals for each cluster'
pVals = {} # CID; [lam,pVal]
for CID in exonDists:
if not len(exonDists[CID]) > 0: #no data in 2kb range.
lam = 'NA'
pVal = 'NA'
else:
lam = cgStats.getLam(exonDists[CID])
pVal = cgStats.getPValExp(predExpression[CID], lam)
pVals[CID] = [
lam, pVal
] #lam gives a good approximation of noise levels in region...
print 'Updating the file'
#update file...
predFile = open(predName, 'r')
newLines = []
for line in predFile:
CID = line.split('\t')[7]
newLine = cg.appendToLine(line, pVals[CID][0], 14)
newLine = cg.appendToLine(newLine, pVals[CID][1], 15)
newLines.append(newLine)
predFile.close()
predFile = open(predName, 'w')
predFile.writelines(newLines)
predFile.close()
def updateReadDensity(tType):
#go through wig each chromosome and check the mature seqs
mainConf = cgConfig.cgConfig('Main.conf')
conf = cgConfig.cgConfig()
organism = conf.conf['organism']
wigFolder = mainConf.conf['wig%s' % organism]
newLines = []
if tType == 'E':
pFileName = conf.conf['resultsExons']
elif tType == 'I':
pFileName = conf.conf['resultsIntrons']
else:
print 'READ UPDATE FAIL'
print ' Updating Read Density:', tType
for wigFileN in cg.recurseDir(wigFolder, end = '.wig'):
#init
chrom = wigFileN.strip().split('.')[-2]
strand = wigFileN.strip().split('.')[-4]
wigFile = open(wigFileN, 'r')
mirFile = open(pFileName, 'r')
print wigFileN
#get rid of header
wigFile.readline()
print ' populating hitmap'
#populate hitmap
wigMap = {}
for line in wigFile:
value = int(line.strip().split('\t')[3].split('.')[0])
if value > 0:
start = int(line.strip().split('\t')[1])
end = int(line.strip().split('\t')[2])
for i in range(start, end):
wigMap[i] = value
wigFile.close()
print ' calculating hits for mature seqs'
#calculate total hits per mature
for line in mirFile:
mTcc = line.strip().split('\t')[1]
mirID = line.strip().split('\t')[0]
if (mTcc.split(':')[0] == chrom) and (mTcc.split(':')[1] == strand):
#if mirID == '26477.30.106643972': print 'Starting Total Count'
highestHit = 0
for i in range(int(mTcc.split(':')[2]), int(mTcc.split(':')[3])):
#if mirID == '26477.30.106643972': print ' ', i
if i in wigMap:
if wigMap[i] > highestHit:
highestHit = wigMap[i]
#if mirID == '26477.30.106643972': print ' ', i, totalHits, wigMap[i]
newLines.append(cg.appendToLine(line, str(highestHit), 11))
mirFile.close()
print 'Writing New File'
#write new results file
outFile = open(pFileName, 'w')
for line in newLines:
outFile.write(line)
outFile.close()
####NOW UPDATE HIGHEST HIT PER CLUSTER####
clusterCount = {}
pFile = open(pFileName, 'r')
for line in pFile:
predictionCount = int(line.strip().split('\t')[11])
CID = line.strip().split('\t')[7]
if CID in clusterCount:
if clusterCount[CID] < predictionCount:
clusterCount[CID] = predictionCount
else:
clusterCount[CID] = predictionCount
pFile.close()
#update the file --> cluster small count
newLines = []
predFile = open(pFileName, 'r')
for line in predFile:
CID = line.strip().split('\t')[7]
numMax = clusterCount[CID]
newLines.append(cg.appendToLine(line, str(numMax), 12))
predFile.close()
#sort newLines by clusterID
sortDict = {}
CIDs = []
for line in newLines:
CID = int(line.strip().split('\t')[7])
if CID not in CIDs:
CIDs.append(CID)
if CID in sortDict:
sortDict[CID].append(line)
else:
sortDict[CID] = [line]
CIDs.sort()
newLines = []
for CID in CIDs:
for line in sortDict[CID]:
newLines.append(line)
#write new File
newFile = open(pFileName, 'w')
for line in newLines:
newFile.write(line)
newFile.close()
def filterOutTargets(resultsFN, centerFN, mismatchFN, targetFN, tranCheck, mPick, cPick, minCenterLevel, inputPosition, updatePosition, outFN):
'''Pick is the range in which you need the values for. 0 is 4bp around, 1 is 6...'''
#make mismatch dict
mmDict = {} # siD: tID : mmVal
f = open(mismatchFN, 'r')
for line in f:
ls = line.strip().split('\t')
sID = int(ls[0])
tID = int(ls[1])
mmVal = int(ls[2 + mPick])
if not sID in mmDict:
mmDict[sID] = {}
mmDict[sID][tID] = mmVal
f.close()
#make center dict
centerDict = {} # sID: tID: centerVal
f = open(centerFN, 'r')
for line in f:
ls = line.strip().split('\t')
sID = int(ls[0])
tID = int(ls[1])
centerVal = float(ls[2 + cPick])
if not sID in centerDict:
centerDict[sID] = {}
centerDict[sID][tID] = centerVal
f.close()
#make transcript target dict
tranVals = {} # tID : tranValue
f = open(targetFN, 'r')
for line in f:
ls = line.strip().split('\t')
tID = int(ls[0])
tranVal = ls[3]
tranVals[tID] = tranVal
#go through and test all the targets.
f = open(resultsFN, 'r')
newLines = []
for line in f:
sID = int(line.strip().split('\t')[0])
targets = line.strip().split('\t')[int(inputPosition)]
targets = targets.strip().split(',')
newTargetList = []
for tID in targets:
##print sID, tID
tID = int(tID)
#Check for inside transcript
if tranCheck:
if tranVals[tID] == '0':
continue
#check mismatches
mmVal = mmDict[sID][tID]
if mmVal == 1:
continue
#check center Expression
centerVal = centerDict[sID][tID]
if centerVal < minCenterLevel:
continue
newTargetList.append(str(tID))
if len(newTargetList) < 1: continue
newTargets = ','.join(newTargetList)
#update newLines
newLines.append(bioLibCG.appendToLine(line, newTargets, int(updatePosition)))
f.close()
#update file
f = open(outFN, 'w')
f.writelines(newLines)
f.close()
def filterOutTargets(resultsFN, centerFN, mismatchFN, targetFN, tranCheck,
mPick, cPick, minCenterLevel, inputPosition,
updatePosition, outFN):
'''Pick is the range in which you need the values for. 0 is 4bp around, 1 is 6...'''
#make mismatch dict
mmDict = {} # siD: tID : mmVal
f = open(mismatchFN, 'r')
for line in f:
ls = line.strip().split('\t')
sID = int(ls[0])
tID = int(ls[1])
mmVal = int(ls[2 + mPick])
if not sID in mmDict:
mmDict[sID] = {}
mmDict[sID][tID] = mmVal
f.close()
#make center dict
centerDict = {} # sID: tID: centerVal
f = open(centerFN, 'r')
for line in f:
ls = line.strip().split('\t')
sID = int(ls[0])
tID = int(ls[1])
centerVal = float(ls[2 + cPick])
if not sID in centerDict:
centerDict[sID] = {}
centerDict[sID][tID] = centerVal
f.close()
#make transcript target dict
tranVals = {} # tID : tranValue
f = open(targetFN, 'r')
for line in f:
ls = line.strip().split('\t')
tID = int(ls[0])
tranVal = ls[3]
tranVals[tID] = tranVal
#go through and test all the targets.
f = open(resultsFN, 'r')
newLines = []
for line in f:
sID = int(line.strip().split('\t')[0])
targets = line.strip().split('\t')[int(inputPosition)]
targets = targets.strip().split(',')
newTargetList = []
for tID in targets:
##print sID, tID
tID = int(tID)
#Check for inside transcript
if tranCheck:
if tranVals[tID] == '0':
continue
#check mismatches
mmVal = mmDict[sID][tID]
if mmVal == 1:
continue
#check center Expression
centerVal = centerDict[sID][tID]
if centerVal < minCenterLevel:
continue
newTargetList.append(str(tID))
if len(newTargetList) < 1: continue
newTargets = ','.join(newTargetList)
#update newLines
newLines.append(
bioLibCG.appendToLine(line, newTargets, int(updatePosition)))
f.close()
#update file
f = open(outFN, 'w')
f.writelines(newLines)
f.close()
def findPeaks(pType, cName = None):
#init
mConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
if pType == 'E':
predName = conf.conf['resultsExonsSorted']
else:
predName = conf.conf['resultsIntronsSorted']
print predName
#make CID:hairpin:peak dictionary
cHairs = getHairpins.getHairpins(predName)
peakDict = {}
for CID in cHairs:
peakDict[CID] = [cHairs[CID],'None']
timer = cg.cgTimer()
timer.start()
#put peaks in memory
print 'Creating peak data'
peaks = {} # chr:peak:value
for CID in cHairs:
chrom, strand, start, end = cg.tccSplit(cHairs[CID])
tcc = cHairs[CID]
#init dictionary
if chrom not in peaks:
peaks[chrom] = {}
if strand not in peaks[chrom]:
peaks[chrom][strand] = {}
#create peaks for tcc and add to peak dictionary
stretch = cgPeaks.stretch(tcc, cName)
stretch.createPeaks()
for peakCoord in stretch.peaks:
peaks[chrom][strand][peakCoord] = 0
print timer.split()
print 'finding best combos'
bestCombos = []
aPass = 0
bPass = 0
cPass = 0
numT = 0
for CID in peakDict:
cgFlag = False
if CID == '538':cgFlag = True
tcc = peakDict[CID][0]
#print tcc
tccPeaks = []
chrom = cg.ss(tcc, ':')[0]
strand = cg.ss(tcc, ':')[1]
start = int(cg.ss(tcc, ':')[2])
end = int(cg.ss(tcc, ':')[3])
#get all peaks
for i in range(start, end + 1):
if i in peaks[chrom][strand]:
#print ' peak added', i
tccPeaks.append(i)
#Calculate parameters...
pairStrings = [] #used to check if pair already added
peakCombos = []
for x in tccPeaks:
#scan a 30 bp range around this point and find the best roof...
pRange = 30
rTcc = cg.makeTcc(chrom, strand, x, x + 1)
#quickly get max value...kinda a long way to do it but whatever
cProfile = stepVectorScan.profileAroundPoint(rTcc, 1, cName, ratio = False)
xval = cProfile[0]
max = xval
highestValueCoord = x
#now make profile for roof...
cProfile = stepVectorScan.profileAroundPoint(rTcc, pRange, cName, ratio = True)
#now get highest stretch length and the rNext coord.
minVal = .80
highest = 0
stretch = 0
startCurrent = None
startFinal = None
endFinal = None
for i in range(1 - pRange, pRange):
if cProfile[i] > minVal:
stretch += 1
if startCurrent == None:
startCurrent = i
else:
if stretch > 0:
if stretch > highest: #stretch ended and was higher than previous
highest = stretch
endFinal = i - 1
startFinal = startCurrent
startCurrent = None
else:
startCurrent = None
stretch = 0
#get +/- 4 value...
val = [1.0, 1.0]
if (startFinal) and (endFinal):
low = startFinal - 4
high = endFinal + 4
if low > (1 - pRange):
if high < pRange:
val[0] = float(cProfile[startFinal - 4])
val[1] = float(cProfile[endFinal + 4])
#fill in other details...
y = 'S'
dist = 'S'
ratio = 'S'
peakCombos.append([tcc,x,y,dist,ratio,max,highest,val])
#print ' ', peakCombos[-1]
#find best combo...
topCombo = None
for combo in peakCombos:
roofLength = combo[6]
dropValue = combo[7][0]
if combo[7][1] > dropValue:
dropValue = combo[7][1]
#print roofLength, dropValue
if 14 < roofLength < 26:
if 0.0 < dropValue < 0.2:
#pick one with rooflength nearest 20:
if topCombo:
if (math.fabs(22 - roofLength)) < (math.fabs(22 - topCombo[6])):
topCombo = combo
else:
topCombo = combo
if topCombo:
peakDict[CID][1] = topCombo
bestCombos.append(topCombo)
print bestCombos[-1]
else:
#print 'None'
pass
print timer.split()
#now update predFile (SLOT 13)
predFile = open(predName, 'r')
newLines = []
for line in predFile:
CID = cg.ss(line)[7]
if peakDict[CID][1] == 'None':
peakInfo = 'None'
else:
peakInfo = '%s:%s:%s:%s:%s:%s' % (str(peakDict[CID][1][1])[-3:], 'S', str(peakDict[CID][1][4]).split('.')[0], peakDict[CID][1][5],peakDict[CID][1][6], peakDict[CID][1][7])
newLines.append(cg.appendToLine(line, peakInfo, 13))
predFile.close()
predFile = open(predName, 'w')
predFile.writelines(newLines)
predFile.close()
