def testmerge(masterDir, parDir):
'''The master directory will contain the merged objects,
the slave directory contains the directories of all the runs
oRNA (master)
aDir (master)
pRuns
--run.00
----oRNA (slave: pRuns/run.00/oRNA)
----aDir
--run.01
'''
mDC = cgDB.dataController(masterDir, cgAlignment.cgAlignment)
id_masterObj = mDC.load()
#recurse through all the runs
masterBN = bioLibCG.getBaseFileName(masterDir)
for slaveDir in bioLibCG.recursePaths(parDir, end = masterBN):
oDC = cgDB.dataController(slaveDir, cgAlignment.cgAlignment)
id_slaveObj = oDC.load()
id_masterObj = cgDB.mergeTwoObjects(id_masterObj, id_slaveObj, cgOriginRNA.OriginRNA)
mDC.commit(id_masterObj)
def filterTargets(oRNADir, aDir, inTranscript, misLevel, centerLevel, minCenterLevel):
if inTranscript == 'True': inTranscript = True
if inTranscript == 'False': inTranscript = False
misLevel, centerLevel, minCenterLevel = int(misLevel), int(centerLevel), float(minCenterLevel)
oDC = cgDB.dataController(oRNADir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for oRNA in id_oRNA.values():
oRNA.filteredTargets = []
for aID in oRNA.targets:
alignment = id_alignment[aID]
#transcriptOverlap
if inTranscript:
if not alignment.transcriptOverlap:
#print 'tOverlap Fail', cgAlignment.pretty#print(alignment)
continue
#misLevel
if alignment.mismatchStatus[misLevel]:
#print 'mismatch Fail', cgAlignment.pretty#print(alignment)
continue
#centerLevel
if alignment.centerExpression[centerLevel] < minCenterLevel:
#print 'expression Fail', cgAlignment.pretty#print(alignment)
continue
oRNA.filteredTargets.append(aID)
oDC.commit(id_oRNA)
def testmerge(masterDir, parDir):
'''The master directory will contain the merged objects,
the slave directory contains the directories of all the runs
oRNA (master)
aDir (master)
pRuns
--run.00
----oRNA (slave: pRuns/run.00/oRNA)
----aDir
--run.01
'''
mDC = cgDB.dataController(masterDir, cgAlignment.cgAlignment)
id_masterObj = mDC.load()
#recurse through all the runs
masterBN = bioLibCG.getBaseFileName(masterDir)
for slaveDir in bioLibCG.recursePaths(parDir, end=masterBN):
oDC = cgDB.dataController(slaveDir, cgAlignment.cgAlignment)
id_slaveObj = oDC.load()
id_masterObj = cgDB.mergeTwoObjects(id_masterObj, id_slaveObj,
cgOriginRNA.OriginRNA)
mDC.commit(id_masterObj)
def plotPairs(oDir, aDir, cName):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for oID, oRNA in id_oRNA.items():
if not oRNA.passedFilter:
continue
for aID in oRNA.filteredTargets:
alignment = id_alignment[aID]
chrom, strand, start, end = bioLibCG.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
print sLen
print oRNA.sequence
print oRNA.tcc
print alignment.tTcc
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
print chrom, strand, start, end
scanRange = bioLibCG.makeTcc(chrom, strand, start, end)
stretch = cgPeaks.stretch(scanRange, cName)
sortedKeys = stretch.profile.keys()
sortedKeys.sort()
if strand == '-1':
sortedKeys.reverse()
xVals = range(1, sLen + 2)
xVals = sortedKeys
yVals = [stretch.profile[x] for x in sortedKeys]
print xVals, len(xVals)
print yVals, len(yVals)
plt.plot(xVals, yVals)
plt.show()
return 0
def plotPairs(oDir, aDir, cName):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for oID, oRNA in id_oRNA.items():
if not oRNA.passedFilter:
continue
for aID in oRNA.filteredTargets:
alignment = id_alignment[aID]
chrom, strand, start, end = bioLibCG.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
print sLen
print oRNA.sequence
print oRNA.tcc
print alignment.tTcc
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
print chrom, strand, start, end
scanRange = bioLibCG.makeTcc(chrom, strand, start, end)
stretch = cgPeaks.stretch(scanRange, cName)
sortedKeys = stretch.profile.keys()
sortedKeys.sort()
if strand == '-1':
sortedKeys.reverse()
xVals = range(1, sLen + 2)
xVals = sortedKeys
yVals = [stretch.profile[x] for x in sortedKeys]
print xVals, len(xVals)
print yVals, len(yVals)
plt.plot(xVals, yVals)
plt.show()
return 0
def markMismatchedPairs(aDir):
#make mismatchDict
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
alignment.mismatchStatus = [False, False, False]
lowRange = range(9,13)
midRange = range(8,14)
highRange = range(7,15)
#check mismatches
for i in lowRange:
if i in alignment.mismatchPositions:
alignment.mismatchStatus[0] = True
break
for i in midRange:
if i in alignment.mismatchPositions:
alignment.mismatchStatus[1] = True
break
for i in highRange:
if i in alignment.mismatchPositions:
alignment.mismatchStatus[2] = True
break
aDC.commit(id_alignment)
def totalSNR(oDir, filterList):
fList = []
f = open(filterList, 'r')
for line in f:
ls = line.strip().split('\t')
fList.append(int(line.strip()))
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
totalRun = 0
totalSim = 0
simsTotals = []
n = 0
for oRNA in id_oRNA.values():
if oRNA.id in fList:
if len(oRNA.filteredTargets) == 0:
continue
totalRun += len(oRNA.filteredTargets)
simsTotal = oRNA.avgNumSimulationTargets * 10
simsTotals.append(simsTotal)
totalSim += oRNA.avgNumSimulationTargets
n +=1
print 'Total Number Targets for my run:', totalRun
print 'Total Number Targets for Simulations:', totalSim
print 'SNR', float(totalRun)/float(totalSim)
print 'stderr(%s)' % n, stdv(simsTotals)/sqrt(10)
def uniqueTargets(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
aID_numHit = {}
uniqueTargets = []
totalTargets = []
for oRNA in id_oRNA.values():
if not oRNA.passedFilter:
continue
for aID in oRNA.filteredTargets:
numHits = aID_numHit.get(aID, 0)
aID_numHit[aID] = numHits + 1
if aID not in uniqueTargets:
uniqueTargets.append(aID)
totalTargets.append(aID)
for aID, numHit in aID_numHit.items():
print aID, numHit
print len(uniqueTargets)
print len(totalTargets)
def loadAlignments2(aDir, alignmentFN):
'''Just added IDs at the beginning to parallel things'''
aDC = cgDB.dataController(aDir, cgAlignment)
id_alignment = {}
f = open(alignmentFN, 'r')
for line in f:
ls = line.strip().split(' ')
id = int(ls[0])
a = cgAlignment(id)
a.sID, a.tID = int(ls[1]), int(ls[2])
a.sStart, a.sEnd = int(ls[3]), int(ls[4])
a.tStart, a.tEnd = int(ls[5]), int(ls[6])
a.sLength, a.tLength = int(ls[7]), int(ls[8])
a.numMismatches = int(ls[9])
try:
a.mismatchPositions = [int(x) for x in ls[10].split(',')]
except IndexError:
a.mismatchPositions = []
id_alignment[id] = a
aDC.commit(id_alignment)
def updateEndContig(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
seq = oRNA.sequence
#5'
cLength5 = 1
for i, letter in enumerate(seq):
if i == 0: continue
if seq[i] == seq[i - 1]:
cLength5 += 1
else:
break
#3'
cLength = 1
revSeq = [x for x in reversed(seq)]
for i, letter in enumerate(revSeq):
if i == 0: continue
if revSeq[i] == revSeq[i - 1]:
cLength += 1
else:
break
highest = cLength5
if cLength > cLength5:
highest = cLength
oRNA.endContigLength = highest
oDC.commit(id_oRNA)
def transcriptSetOverlap(aDir, AS):
AS = bool(AS)
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
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
if AS == True:
degTccs = [cg.convertToAS(x.tcc) for x in id_oRNA.values()]
else:
degTccs = [x.tcc for x in id_oRNA.values()]
#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)
#write new file
for obj in id_oRNA.values():
if AS:
degTcc = cg.convertToAS(obj.tcc)
else:
degTcc = obj.tcc
if degTcc in overlappingDegTccs:
obj.transcriptOverlap = True
else:
obj.transcriptOverlap = False
oRNA_DC.commit(id_oRNA)
def loadAlignments(aDir, alignmentFN):
aDC = cgDB.dataController(aDir, cgAlignment)
id_alignment = {}
i = 0
f = open(alignmentFN, 'r')
for line in f:
ls = line.strip().split(' ')
a = cgAlignment(i)
a.sID, a.tID = int(ls[0]), int(ls[1])
a.sStart, a.sEnd = int(ls[2]), int(ls[3])
a.tStart, a.tEnd = int(ls[4]), int(ls[5])
a.sLength, a.tLength = int(ls[6]), int(ls[7])
a.numMismatches = int(ls[8])
try:
a.mismatchPositions = [int(x) for x in ls[9].split(',')]
except IndexError:
a.mismatchPositions = []
id_alignment[i] = a
i += 1
aDC.commit(id_alignment)
def markMismatchedPairs(aDir):
#make mismatchDict
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
alignment.mismatchStatus = [False, False, False]
lowRange = range(9, 13)
midRange = range(8, 14)
highRange = range(7, 15)
#check mismatches
for i in lowRange:
if i in alignment.mismatchPositions:
alignment.mismatchStatus[0] = True
break
for i in midRange:
if i in alignment.mismatchPositions:
alignment.mismatchStatus[1] = True
break
for i in highRange:
if i in alignment.mismatchPositions:
alignment.mismatchStatus[2] = True
break
aDC.commit(id_alignment)
def uniqueTargets(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
aID_numHit = {}
uniqueTargets = []
totalTargets = []
for oRNA in id_oRNA.values():
if not oRNA.passedFilter:
continue
for aID in oRNA.filteredTargets:
numHits = aID_numHit.get(aID, 0)
aID_numHit[aID] = numHits + 1
if aID not in uniqueTargets:
uniqueTargets.append(aID)
totalTargets.append(aID)
for aID, numHit in aID_numHit.items():
print aID, numHit
print len(uniqueTargets)
print len(totalTargets)
def updateEndContig(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
seq = oRNA.sequence
#5'
cLength5 = 1
for i,letter in enumerate(seq):
if i == 0: continue
if seq[i] == seq[i-1]:
cLength5 += 1
else:
break
#3'
cLength = 1
revSeq = [x for x in reversed(seq)]
for i,letter in enumerate(revSeq):
if i == 0: continue
if revSeq[i] == revSeq[i-1]:
cLength += 1
else:
break
highest = cLength5
if cLength > cLength5:
highest = cLength
oRNA.endContigLength = highest
oDC.commit(id_oRNA)
def loadAlignments(aDir, alignmentFN):
aDC = cgDB.dataController(aDir, cgAlignment)
id_alignment = {}
i = 0
f = open(alignmentFN, 'r')
for line in f:
ls = line.strip().split(' ')
a = cgAlignment(i)
a.sID, a.tID = int(ls[0]), int(ls[1])
a.sStart, a.sEnd = int(ls[2]), int(ls[3])
a.tStart, a.tEnd = int(ls[4]), int(ls[5])
a.sLength, a.tLength = int(ls[6]), int(ls[7])
a.numMismatches = int(ls[8])
try:
a.mismatchPositions = [int(x) for x in ls[9].split(',')]
except IndexError:
a.mismatchPositions = []
id_alignment[i] = a
i += 1
aDC.commit(id_alignment)
def loadAlignments2(aDir, alignmentFN):
'''Just added IDs at the beginning to parallel things'''
aDC = cgDB.dataController(aDir, cgAlignment)
id_alignment = {}
f = open(alignmentFN, 'r')
for line in f:
ls = line.strip().split(' ')
id = int(ls[0])
a = cgAlignment(id)
a.sID, a.tID = int(ls[1]), int(ls[2])
a.sStart, a.sEnd = int(ls[3]), int(ls[4])
a.tStart, a.tEnd = int(ls[5]), int(ls[6])
a.sLength, a.tLength = int(ls[7]), int(ls[8])
a.numMismatches = int(ls[9])
try:
a.mismatchPositions = [int(x) for x in ls[10].split(',')]
except IndexError:
a.mismatchPositions = []
id_alignment[id] = a
aDC.commit(id_alignment)
def transcriptSetOverlap(aDir, AS):
AS = bool(AS)
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
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
if AS == True:
degTccs = [cg.convertToAS(x.tcc) for x in id_oRNA.values()]
else:
degTccs = [x.tcc for x in id_oRNA.values()]
#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)
#write new file
for obj in id_oRNA.values():
if AS:
degTcc = cg.convertToAS(obj.tcc)
else:
degTcc = obj.tcc
if degTcc in overlappingDegTccs:
obj.transcriptOverlap = True
else:
obj.transcriptOverlap = False
oRNA_DC.commit(id_oRNA)
def probeORNA(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
if oRNA.passedFilter:
cgOriginRNA.prettyPrint(oRNA)
def probeORNA(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
if oRNA.passedFilter:
cgOriginRNA.prettyPrint(oRNA)
def transcriptSetOverlapTargets(aDir):
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
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
#create list of unique tccs.
uniqTccs = []
for alignment in id_alignment.values():
chrom, strand, start, end = cg.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
tcc = cg.makeTcc(chrom, strand, start, end)
if tcc not in uniqTccs: uniqTccs.append(tcc)
degTccs = [cg.convertToAS(x) for x in uniqTccs]
#find all overlapping exons/transcripts, then all results sequences that overlap exons
overlappingExons = allExons.transcriptOverlaps(degTccs)
overlappingExonTccs = [x.tcc for x in overlappingExons]
overlappingDegTccs = compare.compareTwoTcc(degTccs, overlappingExonTccs, 1)
#update
for obj in id_alignment.values():
chrom, strand, start, end = cg.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
tcc = cg.makeTcc(chrom, strand, start, end)
degTcc = cg.convertToAS(tcc)
if degTcc in overlappingDegTccs:
obj.transcriptOverlap = True
else:
obj.transcriptOverlap = False
aDC.commit(id_alignment)
def test(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
print id_oRNA[1].targets
id_oRNA[1].targets.append(13)
print id_oRNA[1].targets
print id_oRNA[2].targets
def test(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
print id_oRNA[1].targets
id_oRNA[1].targets.append(13)
print id_oRNA[1].targets
print id_oRNA[2].targets
def probeMicro(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
if oRNA.passedFilter:
print oRNA.id, oRNA.sequence, oRNA.tcc, oRNA.tccs
def probeMicro(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
if oRNA.passedFilter:
print oRNA.id, oRNA.sequence, oRNA.tcc, oRNA.tccs
def transcriptSetOverlapTargets(aDir):
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
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
#create list of unique tccs.
uniqTccs = []
for alignment in id_alignment.values():
chrom, strand, start, end = cg.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
tcc = cg.makeTcc(chrom, strand, start, end)
if tcc not in uniqTccs: uniqTccs.append(tcc)
degTccs = [cg.convertToAS(x) for x in uniqTccs]
#find all overlapping exons/transcripts, then all results sequences that overlap exons
overlappingExons = allExons.transcriptOverlaps(degTccs)
overlappingExonTccs = [x.tcc for x in overlappingExons]
overlappingDegTccs = compare.compareTwoTcc(degTccs, overlappingExonTccs, 1)
#update
for obj in id_alignment.values():
chrom, strand, start, end = cg.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
tcc = cg.makeTcc(chrom, strand, start, end)
degTcc = cg.convertToAS(tcc)
if degTcc in overlappingDegTccs:
obj.transcriptOverlap = True
else:
obj.transcriptOverlap = False
aDC.commit(id_alignment)
def updateEntropy(aDir):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
for obj in id_oRNA.values():
obj.entropy = getEntropy(obj.sequence)
oRNA_DC.commit(id_oRNA)
def pPrint(aDir):
aDC = cgDB.dataController(aDir, cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
attName_att = alignment.__dict__
attVals = [attName_att['id'], attName_att['tID'], attName_att['tTcc']]
attVals = [str(x) for x in attVals]
print '\t'.join(attVals)
def pPrint(aDir):
aDC = cgDB.dataController(aDir, cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
attName_att = alignment.__dict__
attVals = [attName_att['id'], attName_att['tID'], attName_att['tTcc']]
attVals = [str(x) for x in attVals]
print '\t'.join(attVals)
def updateEntropy(aDir):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
for obj in id_oRNA.values():
obj.entropy = getEntropy(obj.sequence)
oRNA_DC.commit(id_oRNA)
def probeAlignments(aDir):
probePairs = [[6, 35934]]
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
for sID, tID in probePairs:
if alignment.sID == sID and alignment.tID == tID:
print alignment.id, alignment.sID, alignment.tID, alignment.centerExpression, alignment.mismatchStatus, alignment.numMismatches, alignment.transcriptOverlap
def updateTargetIDs(oDir, aDir):
#load the data
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
oRNA_DC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
#clear targets that are there.
for oRNA in id_oRNA.values():
oRNA.targets = []
#update the targets for oRNAs
for alignment in id_alignment.values():
try:
id_oRNA[alignment.sID].targets.append(alignment.id)
except KeyError:
print 'oRNA key missing', alignment.sID, alignment.id
#save
oRNA_DC.commit(id_oRNA)
def probeAlignments(aDir):
probePairs = [[6, 35934]]
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
for sID, tID in probePairs:
if alignment.sID == sID and alignment.tID == tID:
print alignment.id, alignment.sID, alignment.tID, alignment.centerExpression, alignment.mismatchStatus, alignment.numMismatches, alignment.transcriptOverlap
def updateTargetIDs(oDir, aDir):
#load the data
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
oRNA_DC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
#clear targets that are there.
for oRNA in id_oRNA.values():
oRNA.targets = []
#update the targets for oRNAs
for alignment in id_alignment.values():
try:
id_oRNA[alignment.sID].targets.append(alignment.id)
except KeyError:
print 'oRNA key missing', alignment.sID, alignment.id
#save
oRNA_DC.commit(id_oRNA)
def updateID(aDir, peakFN):
'''This fxn will initialize the database objects...'''
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = {}
f = open(peakFN, 'r')
i = 0
for line in f:
id_oRNA[i] = cgOriginRNA.OriginRNA(i)
i += 1
oRNA_DC.commit(id_oRNA)
def updateID(aDir, peakFN):
'''This fxn will initialize the database objects...'''
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = {}
f = open(peakFN, 'r')
i = 0
for line in f:
id_oRNA[i] = cgOriginRNA.OriginRNA(i)
i += 1
oRNA_DC.commit(id_oRNA)
def updateIDFromQuery(aDir, queryFN):
'''This fxn will initialize the database objects...'''
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = {}
f = open(queryFN, 'r')
for line in f:
ls = line.strip().split('\t')
id = int(ls[0])
id_oRNA[id] = cgOriginRNA.OriginRNA(id)
oRNA_DC.commit(id_oRNA)
def updateIDFromQuery(aDir, queryFN):
'''This fxn will initialize the database objects...'''
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = {}
f = open(queryFN, 'r')
for line in f:
ls = line.strip().split('\t')
id = int(ls[0])
id_oRNA[id] = cgOriginRNA.OriginRNA(id)
oRNA_DC.commit(id_oRNA)
def updateSmallExpression(aDir, cName):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
for id, oRNA in id_oRNA.items():
stretch = cgPeaks.stretch(oRNA.tcc, cName) #this stretch contains values for small library...
highValue = stretch.getHighestLevel()
oRNA.eLevel = highValue
oRNA_DC.commit(id_oRNA)
def filterORNA(oDir,
maxEndContig,
maxTotalContig,
minSNR,
minNumTargets,
keepDuplicates=False):
if keepDuplicates == 'True': keepDuplicates = True
if keepDuplicates == 'False': keepDuplicates = False
maxEndContig, maxTotalContig = int(maxEndContig), int(maxTotalContig)
minNumTargets = int(minNumTargets)
minSNR = float(minSNR)
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
oRNA.passedFilter = True
if len(oRNA.filteredTargets) < minNumTargets:
oRNA.passedFilter = False
cgOriginRNA.prettyPrint(oRNA, 'numTargets')
continue
if oRNA.endContigLength > maxEndContig:
cgOriginRNA.prettyPrint(oRNA, 'endContig')
oRNA.passedFilter = False
continue
if oRNA.totalContigLength > maxTotalContig:
cgOriginRNA.prettyPrint(oRNA, 'totalContig')
oRNA.passedFilter = False
continue
if oRNA.snr < minSNR:
cgOriginRNA.prettyPrint(oRNA, 'SNR fail')
oRNA.passedFilter = False
continue
if not keepDuplicates:
if oRNA.sequenceDuplicate:
cgOriginRNA.prettyPrint(oRNA, 'Duplicate Fail')
oRNA.passedFilter = False
continue
print 'PASSED:', oRNA.id, ','.join(
str(x) for x in oRNA.filteredTargets
), oRNA.entropy, oRNA.avgNumSimulationTargets, oRNA.snr, oRNA.endContigLength, oRNA.sequence
oDC.commit(id_oRNA)
def updateSmallExpression(aDir, cName):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
for id, oRNA in id_oRNA.items():
stretch = cgPeaks.stretch(
oRNA.tcc,
cName) #this stretch contains values for small library...
highValue = stretch.getHighestLevel()
oRNA.eLevel = highValue
oRNA_DC.commit(id_oRNA)
def updateSequence(aDir, seqFN):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
f = open(seqFN, 'r')
i = 0
for line in f:
ls = line.strip().split('\t')
seq = ls[0]
id_oRNA[i].sequence = seq
i += 1
oRNA_DC.commit(id_oRNA)
def updateTcc(aDir, tccFN):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
f = open(tccFN, 'r')
i = 0
for line in f:
ls = line.strip().split('\t')
tcc = ls[0]
id_oRNA[i].tcc = tcc
i += 1
oRNA_DC.commit(id_oRNA)
def updateSequence(aDir, seqFN):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
f = open(seqFN, 'r')
i = 0
for line in f:
ls = line.strip().split('\t')
seq = ls[0]
id_oRNA[i].sequence = seq
i += 1
oRNA_DC.commit(id_oRNA)
def updateTcc(aDir, tccFN):
oRNA_DC = cgDB.dataController(aDir, cgOriginRNA.OriginRNA)
id_oRNA = oRNA_DC.load()
f = open(tccFN, 'r')
i = 0
for line in f:
ls = line.strip().split('\t')
tcc = ls[0]
id_oRNA[i].tcc = tcc
i += 1
oRNA_DC.commit(id_oRNA)
def filterTargets(oRNADir, aDir, inTranscript, misLevel, centerLevel,
minCenterLevel):
if inTranscript == 'True': inTranscript = True
if inTranscript == 'False': inTranscript = False
misLevel, centerLevel, minCenterLevel = int(misLevel), int(
centerLevel), float(minCenterLevel)
oDC = cgDB.dataController(oRNADir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for oRNA in id_oRNA.values():
oRNA.filteredTargets = []
for aID in oRNA.targets:
alignment = id_alignment[aID]
#transcriptOverlap
if inTranscript:
if not alignment.transcriptOverlap:
#print 'tOverlap Fail', cgAlignment.pretty#print(alignment)
continue
#misLevel
if alignment.mismatchStatus[misLevel]:
#print 'mismatch Fail', cgAlignment.pretty#print(alignment)
continue
#centerLevel
if alignment.centerExpression[centerLevel] < minCenterLevel:
#print 'expression Fail', cgAlignment.pretty#print(alignment)
continue
oRNA.filteredTargets.append(aID)
oDC.commit(id_oRNA)
def getSeqs(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for id, oRNA in id_oRNA.items():
if oRNA.sequenceDuplicate:
continue
if oRNA.totalContigLength > 6:
continue
if oRNA.endContigLength > 6:
continue
print "%s" % id
def updateAvgNumTargets(oDir):
oID_numTargets = {}
for i in range(0,10):
print i
simDirRNA = '/home/chrisgre/scripts/simulations/simsk50Filtered/simulation.%s/oRNA' % i
oDC = cgDB.dataController(simDirRNA, cgOriginRNA.OriginRNA)
id_sRNA = oDC.load()
for id, sRNA in id_sRNA.items():
currTargets = oID_numTargets.get(id, 0)
oID_numTargets[id] = currTargets + len(sRNA.filteredTargets)
#now save it
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
totalNum = oID_numTargets.get(oRNA.id, 0)
avgNum = float(totalNum)/float(10.0)
oRNA.avgNumSimulationTargets = avgNum
oDC.commit(id_oRNA)
def quickScript(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
failedList = []
for id, oRNA in id_oRNA.items():
if oRNA.endContigLength > 7 or oRNA.totalContigLength > 7:
failedList.append(id)
for id in failedList:
print id
def getSeqs(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for id, oRNA in id_oRNA.items():
if oRNA.sequenceDuplicate:
continue
if oRNA.totalContigLength > 6:
continue
if oRNA.endContigLength > 6:
continue
print '%s' % id
def updateSNR(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
actualNum = len(oRNA.filteredTargets)
avgNum = oRNA.avgNumSimulationTargets
if avgNum == 0: avgNum = .01
SNR = float(actualNum) / avgNum
oRNA.snr = SNR
oDC.commit(id_oRNA)
def updateAvgNumTargets(oDir):
oID_numTargets = {}
for i in range(0, 10):
print i
simDirRNA = '/home/chrisgre/scripts/simulations/simsk50Filtered/simulation.%s/oRNA' % i
oDC = cgDB.dataController(simDirRNA, cgOriginRNA.OriginRNA)
id_sRNA = oDC.load()
for id, sRNA in id_sRNA.items():
currTargets = oID_numTargets.get(id, 0)
oID_numTargets[id] = currTargets + len(sRNA.filteredTargets)
#now save it
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
totalNum = oID_numTargets.get(oRNA.id, 0)
avgNum = float(totalNum) / float(10.0)
oRNA.avgNumSimulationTargets = avgNum
oDC.commit(id_oRNA)
def updateSNR(oDir):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
actualNum = len(oRNA.filteredTargets)
avgNum = oRNA.avgNumSimulationTargets
if avgNum == 0: avgNum = .01
SNR = float(actualNum)/avgNum
oRNA.snr = SNR
oDC.commit(id_oRNA)
def markCenterExpression(aDir, cName):
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
alignment.centerExpression = [0.0, 0.0, 0.0]
chrom, strand, start, end = bioLibCG.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
scanRange = bioLibCG.makeTcc(chrom, strand, start, end)
stretch = cgPeaks.stretch(scanRange, cName)
expressionSum = stretch.getSumOfLevels()
sortedKeys = stretch.profile.keys()
sortedKeys.sort()
if strand == '-1':
sortedKeys.reverse()
if expressionSum != 0:
sum = 0.0
for key in sortedKeys[8:12]:
sum += stretch.profile[key]
alignment.centerExpression[0] = sum/expressionSum
sum = 0.0
for key in sortedKeys[7:13]:
sum += stretch.profile[key]
alignment.centerExpression[1] = sum/expressionSum
sum = 0.0
for key in sortedKeys[6:14]:
sum += stretch.profile[key]
alignment.centerExpression[2] = sum/expressionSum
aDC.commit(id_alignment)
def markCenterExpression(aDir, cName):
aDC = cgDB.dataController(aDir, cgAlignment.cgAlignment)
id_alignment = aDC.load()
for alignment in id_alignment.values():
alignment.centerExpression = [0.0, 0.0, 0.0]
chrom, strand, start, end = bioLibCG.tccSplit(alignment.tTcc)
offset = alignment.tStart
sLen = alignment.sLength
if strand == '1':
start = start - 19 + offset
end = start + sLen
else:
end = end + 19 - offset
start = end - sLen
scanRange = bioLibCG.makeTcc(chrom, strand, start, end)
stretch = cgPeaks.stretch(scanRange, cName)
expressionSum = stretch.getSumOfLevels()
sortedKeys = stretch.profile.keys()
sortedKeys.sort()
if strand == '-1':
sortedKeys.reverse()
if expressionSum != 0:
sum = 0.0
for key in sortedKeys[8:12]:
sum += stretch.profile[key]
alignment.centerExpression[0] = sum / expressionSum
sum = 0.0
for key in sortedKeys[7:13]:
sum += stretch.profile[key]
alignment.centerExpression[1] = sum / expressionSum
sum = 0.0
for key in sortedKeys[6:14]:
sum += stretch.profile[key]
alignment.centerExpression[2] = sum / expressionSum
aDC.commit(id_alignment)
def signalHisto(oDir, title = 'SNR'):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
histVals = []
for oRNA in id_oRNA.values():
if not oRNA.passedFilter:
continue
histVals.append(math.log(oRNA.snr, 2))
plt.hist(histVals, 30, facecolor='b', alpha = .75)
plt.axis([-4,10,0,30])
plt.title('%s' % title)
plt.xlabel('log2(Signal/Noise)')
plt.ylabel('Number of Origin RNAs')
plt.show()
def signalHisto(oDir, title='SNR'):
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
histVals = []
for oRNA in id_oRNA.values():
if not oRNA.passedFilter:
continue
histVals.append(math.log(oRNA.snr, 2))
plt.hist(histVals, 30, facecolor='b', alpha=.75)
plt.axis([-4, 10, 0, 30])
plt.title('%s' % title)
plt.xlabel('log2(Signal/Noise)')
plt.ylabel('Number of Origin RNAs')
plt.show()
def filterORNA(oDir, maxEndContig, maxTotalContig, minSNR, minNumTargets, keepDuplicates = False):
if keepDuplicates == 'True': keepDuplicates = True
if keepDuplicates == 'False': keepDuplicates = False
maxEndContig, maxTotalContig = int(maxEndContig), int(maxTotalContig)
minNumTargets = int(minNumTargets)
minSNR = float(minSNR)
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
for oRNA in id_oRNA.values():
oRNA.passedFilter = True
if len(oRNA.filteredTargets) < minNumTargets:
oRNA.passedFilter = False
cgOriginRNA.prettyPrint(oRNA, 'numTargets')
continue
if oRNA.endContigLength > maxEndContig:
cgOriginRNA.prettyPrint(oRNA, 'endContig')
oRNA.passedFilter = False
continue
if oRNA.totalContigLength > maxTotalContig:
cgOriginRNA.prettyPrint(oRNA, 'totalContig')
oRNA.passedFilter = False
continue
if oRNA.snr < minSNR:
cgOriginRNA.prettyPrint(oRNA, 'SNR fail')
oRNA.passedFilter = False
continue
if not keepDuplicates:
if oRNA.sequenceDuplicate:
cgOriginRNA.prettyPrint(oRNA, 'Duplicate Fail')
oRNA.passedFilter = False
continue
print 'PASSED:', oRNA.id, ','.join(str(x) for x in oRNA.filteredTargets), oRNA.entropy, oRNA.avgNumSimulationTargets, oRNA.snr, oRNA.endContigLength, oRNA.sequence
oDC.commit(id_oRNA)
