def smallAnalyze(inFile = None):
#Caste and defaults
if not inFile:
print 'need inFile'
return 1
# start Timer
timer = cg.cgTimer()
timer.start()
# for every id in outfile, count how many matches there are
countDict = {}
countFile = open(inFile, 'r')
for line in countFile:
(id, library) = (line.strip().split(':')[1], line.strip().split(':')[0])
if id not in countDict:
countDict[id] = {}
else:
if library not in countDict[id]:
countDict[id][library] = 1
else:
countDict[id][library] = countDict[id][library] + 1
#print 'Time for counting lib hits: ', timer.split()
sortList = []
for id in countDict:
sortList.append(id)
sortList.sort()
for id in sortList:
print '%s' % id
for lib in countDict[id]:
print '%s\t%s' % (lib, countDict[id][lib])
def markCenterExpression(aFN, wigDir, rn = None, tn = None):
extend = 25
timer = bioLibCG.cgTimer()
timer.start()
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['centerExpression', 'tTcc', 'tStart', 'sLength', 'tELevel'], [rn, tn])
#load expression of degradome
wigDict = cgWig.loadWigDict(wigDir)
for aID in aNX.centerExpression:
aNX.centerExpression[aID] = [0.0, 0.0, 0.0]
chrom, strand, start, end = bioLibCG.tccSplit(aNX.tTcc[aID])
offset = aNX.tStart[aID]
sLen = aNX.sLength[aID]
if strand == '1':
start = start - extend + offset
end = start + sLen
else:
end = end + extend - offset
start = end - sLen
scanRange = bioLibCG.makeTcc(chrom, strand, start, end)
stretch = cgWig.getExpressionProfile(scanRange, wigDict)
#make sure peak is in the small range
peakLevel = aNX.tELevel[aID]
peakInRange = (peakLevel in stretch.values())
expressionSum = sum(stretch.values())
sortedKeys = stretch.keys()
sortedKeys.sort()
if strand == '-1':
sortedKeys.reverse()
if expressionSum != 0 and peakInRange:
sumE = 0.0
for key in sortedKeys[8:12]:
sumE += stretch[key]
aNX.centerExpression[aID][0] = sumE/expressionSum
sumE = 0.0
for key in sortedKeys[7:13]:
sumE += stretch[key]
aNX.centerExpression[aID][1] = sumE/expressionSum
sumE = 0.0
for key in sortedKeys[6:14]:
sumE += stretch[key]
aNX.centerExpression[aID][2] = sumE/expressionSum
aNX.save()
def alignSeqs(seqsFN, dbName, wordSize, outFN, maxNumMismatches, sendExitSignal = False):
maxNumMismatches = int(maxNumMismatches)
sendExitSignal = bool(sendExitSignal)
timer = bioLibCG.cgTimer()
timer.start()
#put seqs in cgSeq object, align
wName = dbName + '.wDB'
sName = dbName + '.sDB'
wordSize = int(wordSize)
#load dbs
#print 'loading Sequence Database'
sDB = cgAlign.loadSequenceDatabase(sName)
print timer.split()
#print 'loading Word Database'
wDB = cgAlign.loadWordDatabase(wName)
print timer.split()
#align each seq
f = open(seqsFN, 'r')
fOut = open(outFN, 'w')
for line in f:
qSeq = cgAlign.cgSeq(line.strip().split('\t')[0], line.strip().split('\t')[1])
#write out the alignments
cgAlign.alignQuery(qSeq, wDB, sDB, wordSize, maxNumMismatches, fOut)
f.close()
fOut.close()
print timer.split()
if sendExitSignal:
cgExit.sendExitSignal(seqsFN)
def scanVectorsFile(fN, tccList):
'''Given tcc list --> scan wig files and return coord:value...
'''
timer = cg.cgTimer()
timer.start()
coordDict = {} # tcc: [list values]
for tcc in tccList:
chrom, strand, tccStart, tccEnd = cg.tccSplit(tcc)
#goto correct line in index
fIndex = cgIndex.lineIndex(fN, header = True) #!!!there actually is a header...have to deal with this...
fIndex.passCheckFunction(cgIndex.wigCheckFunction)
fIndex.binarySearch(tcc) #places file pointer at beginning of tcc as beginning
stop = False
for line in fIndex.file:
#print 'Line:', line.strip()
lBeg = int(cg.ss(line)[1])
lEnd = int(cg.ss(line)[2])
lValue = int(cg.ss(line)[3].split('.')[0])
if tccStart > lBeg:
lBeg = tccStart
if tccEnd < lEnd:
lEnd = tccEnd
stop = True
#print timer.split()
for i in range(lBeg, lEnd):
coordDict[i] = lValue
if stop: break
#fIndex.close()
return coordDict
def updatePolySeqs(mFN, readsFN, alignFN):
tim = bioLibCG.cgTimer()
tim.start()
variousAs = ["A" * x for x in range(1,20)]
variousGs = ["G" * x for x in range(1,20)]
variousTs = ["T" * x for x in range(1,20)]
variousCs = ["C" * x for x in range(1,20)]
letter_variousLetters = [ ("A", variousAs),
("G", variousGs),
("T", variousTs),
("C", variousCs)]
checkRange = range(1,8)
NX = cgNexusFlat.Nexus(mFN, miR)
NX.load(['sequence', 'polySeqs'])
#print 'load micro', tim.split()
reads = cgNexusFlat.quickTable(('read','string', '.', 1))
rNX = cgNexusFlat.Nexus(readsFN, reads)
rNX.load(['read'])
#print 'load reads', tim.split()
aNX = cgNexusFlat.Nexus(alignFN, cgAlignment)
aNX.load(['sID', 'tID'])
#print 'load alignments', tim.split()
for id in aNX.ids:
theRead = rNX.read[aNX.sID[id]]
mID = aNX.tID[id]
microSeq = NX.sequence[mID]
#may be a read for expression, but wont count...
if theRead in microSeq: continue
#just for expression
if microSeq == theRead:
print tabIt(microSeq, theRead, 0, 0, "N")
#first check full
elif microSeq in theRead and (len(theRead) != len(microSeq)):
tail = theRead.split(microSeq)[1]
for let, variousLetters in letter_variousLetters:
if tail in variousLetters:
print tabIt(microSeq, theRead, 0, len(tail), let)
#now check trimmed (cant do [:-0])
else:
for i in checkRange:
if microSeq[:-i] in theRead and (len(theRead) != len(microSeq[:-i])):
tail = theRead.split(microSeq[:-i])[1]
for let, variousLetters in letter_variousLetters:
if tail in variousLetters:
print tabIt(microSeq, theRead, i, len(tail), let)
print "TRIMMED"
break #dont trim after the first trimmed one works
def scanVectorsFile(fN, tccList):
'''Given tcc list --> scan wig files and return coord:value...
'''
timer = cg.cgTimer()
timer.start()
coordDict = {} # tcc: [list values]
for tcc in tccList:
chrom, strand, tccStart, tccEnd = cg.tccSplit(tcc)
#goto correct line in index
fIndex = cgIndex.lineIndex(fN, header = True) #!!!there actually is a header...have to deal with this...
fIndex.passCheckFunction(cgIndex.wigCheckFunction)
fIndex.binarySearch(tcc) #places file pointer at beginning of tcc as beginning
stop = False
for line in fIndex.file:
#print 'Line:', line.strip()
lBeg = int(cg.ss(line)[1])
lEnd = int(cg.ss(line)[2])
lValue = int(cg.ss(line)[3].split('.')[0])
if tccStart > lBeg:
lBeg = tccStart
if tccEnd < lEnd:
lEnd = tccEnd
stop = True
#print timer.split()
for i in range(lBeg, lEnd):
coordDict[i] = lValue
if stop: break
#fIndex.close()
return coordDict
def markCenterExpression(aFN, wigDir, rn=None, tn=None):
extend = 25
timer = bioLibCG.cgTimer()
timer.start()
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['centerExpression', 'tTcc', 'tStart', 'sLength', 'tELevel'],
[rn, tn])
#load expression of degradome
wigDict = cgWig.loadWigDict(wigDir)
for aID in aNX.centerExpression:
aNX.centerExpression[aID] = [0.0, 0.0, 0.0]
chrom, strand, start, end = bioLibCG.tccSplit(aNX.tTcc[aID])
offset = aNX.tStart[aID]
sLen = aNX.sLength[aID]
if strand == '1':
start = start - extend + offset
end = start + sLen
else:
end = end + extend - offset
start = end - sLen
scanRange = bioLibCG.makeTcc(chrom, strand, start, end)
stretch = cgWig.getExpressionProfile(scanRange, wigDict)
#make sure peak is in the small range
peakLevel = aNX.tELevel[aID]
peakInRange = (peakLevel in stretch.values())
expressionSum = sum(stretch.values())
sortedKeys = stretch.keys()
sortedKeys.sort()
if strand == '-1':
sortedKeys.reverse()
if expressionSum != 0 and peakInRange:
sumE = 0.0
for key in sortedKeys[8:12]:
sumE += stretch[key]
aNX.centerExpression[aID][0] = sumE / expressionSum
sumE = 0.0
for key in sortedKeys[7:13]:
sumE += stretch[key]
aNX.centerExpression[aID][1] = sumE / expressionSum
sumE = 0.0
for key in sortedKeys[6:14]:
sumE += stretch[key]
aNX.centerExpression[aID][2] = sumE / expressionSum
aNX.save()
def alignSeqs(seqsFN,
dbName,
wordSize,
outFN,
maxNumMismatches,
sendExitSignal=False):
maxNumMismatches = int(maxNumMismatches)
sendExitSignal = bool(sendExitSignal)
timer = bioLibCG.cgTimer()
timer.start()
#put seqs in cgSeq object, align
wName = dbName + '.wDB'
sName = dbName + '.sDB'
wordSize = int(wordSize)
#load dbs
#print 'loading Sequence Database'
sDB = cgAlign.loadSequenceDatabase(sName)
print timer.split()
#print 'loading Word Database'
wDB = cgAlign.loadWordDatabase(wName)
print timer.split()
#align each seq
f = open(seqsFN, 'r')
fOut = open(outFN, 'w')
for line in f:
qSeq = cgAlign.cgSeq(line.strip().split('\t')[0],
line.strip().split('\t')[1])
#write out the alignments
cgAlign.alignQuery(qSeq, wDB, sDB, wordSize, maxNumMismatches, fOut)
f.close()
fOut.close()
print timer.split()
if sendExitSignal:
cgExit.sendExitSignal(seqsFN)
def smallAnalyze(inFile=None):
#Caste and defaults
if not inFile:
print 'need inFile'
return 1
# start Timer
timer = cg.cgTimer()
timer.start()
# for every id in outfile, count how many matches there are
countDict = {}
countFile = open(inFile, 'r')
for line in countFile:
(id, library) = (line.strip().split(':')[1],
line.strip().split(':')[0])
if id not in countDict:
countDict[id] = {}
else:
if library not in countDict[id]:
countDict[id][library] = 1
else:
countDict[id][library] = countDict[id][library] + 1
#print 'Time for counting lib hits: ', timer.split()
sortList = []
for id in countDict:
sortList.append(id)
sortList.sort()
for id in sortList:
print '%s' % id
for lib in countDict[id]:
print '%s\t%s' % (lib, countDict[id][lib])
def scanVectorsSingleCoord(tccList, cName):
'''Given tcc list --> scan wig files and coord:value...
'''
conf = c.getConfig(cName)
org = conf.conf['organism']
mConf = c.getConfig('Main.conf')
wigDir = mConf.conf['wig%s' % org]
timer = cg.cgTimer()
timer.start()
coordDict = {} # tcc: [list values]
for tcc in tccList:
theSplit = ss(tcc, ':')
chrom, strand, tccStart, tccEnd = theSplit[0], theSplit[1], int(
theSplit[2]), int(theSplit[3])
#goto correct fild, correct line in index
fN = wigDir + '/Merge.%s.%s.wig.%s.wig' % (org.lower(), strand, chrom)
fNindex = wigDir + '/Merge.%s.%s.wig.%s.wig.index' % (org.lower(),
strand, chrom)
#print timer.split()
#get line in index file
iFile = open(fNindex, 'r')
startByte = 'None'
for line in iFile:
beg = int(cg.ss(line)[1])
end = int(cg.ss(line)[2])
if beg <= tccStart < end:
startByte = int(cg.ss(line)[0])
#print 'INDEX', line.strip()
break
iFile.close()
#print timer.split()
#grab value
f = open(fN, 'r')
f.seek(startByte, 0)
stop = False
for line in f:
#print 'Line:', line.strip()
lBeg = int(cg.ss(line)[1])
lEnd = int(cg.ss(line)[2])
lValue = int(cg.ss(line)[3].split('.')[0])
if tccStart > lBeg:
lBeg = tccStart
if tccEnd < lEnd:
lEnd = tccEnd
stop = True
#print timer.split()
for i in range(lBeg, lEnd):
coordDict[i] = lValue
if stop: break
f.close()
return coordDict
def scanVectorsHist(tccList, cName):
'''Given tcc list --> scan wig files and get histogram values
can be modified to do single/total values...
THIS USES INDEXES!!! = BAD...'''
conf = c.getConfig(cName)
org = conf.conf['organism']
mConf = c.getConfig('Main.conf')
wigDir = mConf.conf['wig%s' % org]
timer = cg.cgTimer()
timer.start()
histDict = {} # tcc: [list values]
for tcc in tccList:
theSplit = ss(tcc, ':')
chrom, strand, tccStart, tccEnd = theSplit[0], theSplit[1], int(
theSplit[2]), int(theSplit[3])
#goto correct fild, correct line in index
fN = wigDir + '/Merge.%s.%s.wig.%s.wig' % (org.lower(), strand, chrom)
fNindex = wigDir + '/Merge.%s.%s.wig.%s.wig.index' % (org.lower(),
strand, chrom)
#print timer.split()
#get line in index file
iFile = open(fNindex, 'r')
startByte = 'None'
for line in iFile:
beg = int(cg.ss(line)[1])
end = int(cg.ss(line)[2])
if beg <= tccStart < end:
startByte = int(cg.ss(line)[0])
#print 'INDEX', line.strip()
break
iFile.close()
#print timer.split()
#grab value
f = open(fN, 'r')
f.seek(startByte, 0)
stop = False
for line in f:
#print 'Line:', line.strip()
lBeg = int(cg.ss(line)[1])
lEnd = int(cg.ss(line)[2])
lValue = int(cg.ss(line)[3].split('.')[0])
if tccStart > lBeg:
lBeg = tccStart
if tccEnd < lEnd:
lEnd = tccEnd
stop = True
#print timer.split()
for i in range(lBeg, lEnd):
try:
histDict[tcc].append(lValue)
except KeyError: #just for zero...so you don't have to if every time...
histDict[tcc] = [lValue]
if stop: break
f.close()
#print timer.split()
return histDict
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 scanVectorsSingleCoord(tccList, cName):
'''Given tcc list --> scan wig files and coord:value...
'''
conf = c.getConfig(cName)
org = conf.conf['organism']
mConf = c.getConfig('Main.conf')
wigDir = mConf.conf['wig%s' % org]
timer = cg.cgTimer()
timer.start()
coordDict = {} # tcc: [list values]
for tcc in tccList:
theSplit = ss(tcc, ':')
chrom, strand, tccStart, tccEnd = theSplit[0], theSplit[1],int(theSplit[2]),int(theSplit[3])
#goto correct fild, correct line in index
fN = wigDir + '/Merge.%s.%s.wig.%s.wig' % (org.lower(),strand,chrom)
fNindex = wigDir + '/Merge.%s.%s.wig.%s.wig.index' % (org.lower(),strand,chrom)
#print timer.split()
#get line in index file
iFile = open(fNindex, 'r')
startByte = 'None'
for line in iFile:
beg = int(cg.ss(line)[1])
end = int(cg.ss(line)[2])
if beg <= tccStart < end:
startByte = int(cg.ss(line)[0])
#print 'INDEX', line.strip()
break
iFile.close()
#print timer.split()
#grab value
f = open(fN, 'r')
f.seek(startByte, 0)
stop = False
for line in f:
#print 'Line:', line.strip()
lBeg = int(cg.ss(line)[1])
lEnd = int(cg.ss(line)[2])
lValue = int(cg.ss(line)[3].split('.')[0])
if tccStart > lBeg:
lBeg = tccStart
if tccEnd < lEnd:
lEnd = tccEnd
stop = True
#print timer.split()
for i in range(lBeg, lEnd):
coordDict[i] = lValue
if stop: break
f.close()
return coordDict
#puts final hits into clusters... ##Clusters are based off of overlapping neighbors, if you have an overlapping neighbor than you are part of that cluster. import bioLibCG as cg import subprocess import compareData as compare import cgConfig #Start Timer timer = cg.cgTimer() timer.start() #Get list of mature tccs conf = cgConfig.returnConfDict() finalMirFileName = '/u/home8/gxxiao/chrisgre/projects/PipeRuns/LanderHuman/out/LanderHuman-s3k8b17.ALL.FINAL.mirs.tsv' finalMirFileName = conf['resultsRaw'] matureTccs = compare.tccFileToList(finalMirFileName, 1) # list of all mature micro in tcc print 'List getting', timer.split() #make connections dict matureConnections = compare.makeConnectionsDict(matureTccs) print 'Make connections:', timer.split() #Now have to define Clusters... clusters = [] addedList = [] #I don't think python passes by reference? also I think this function is in the middle because it uses a global variable :P def createClusters(item = None, mode = None):
def defineClusters(cName = None):
#Start Timer
timer = cg.cgTimer()
timer.start()
#Get list of mature tccs
conf = cgConfig.getConfig(cName) #passed or default
finalMirFileName = conf.conf['resultsRaw']
matureTccs = compare.tccFileToList(finalMirFileName, 1) # list of all mature micro in tcc
print 'List getting', timer.split()
#make connections dict
matureConnections = compare.makeConnectionsDict(matureTccs)
print 'Make connections:', timer.split()
#Now have to define Clusters...
clusters = []
addedList = []
#I don't think python passes by reference? also I think this function is in the middle because it uses a global variable :P
def createClusters(item = None, mode = None):
if item in addedList:
return 0
elif mode == "top":
clusters.append([item])
addedList.append(item) ##creates new cluster with the item already stored in it
for connectedItem in matureConnections[item]:
createClusters(connectedItem, "neighbor")
elif mode == "neighbor":
clusters[-1].append(item) #add this item to the last cluster created
addedList.append(item)
for connectedItem in matureConnections[item]:
createClusters(connectedItem, "neighbor")
for tcc in matureTccs:
createClusters(tcc, "top")
print 'Make Clusters', timer.split()
#Sort Clusters.
sortedClusters = []
for cluster in clusters:
sortedClusters.append(cg.sortTccList(cluster))
print 'Sort Clusters:', timer.split()
#Output sorted cluster file
clusterFileName = conf.conf['sortedClusters']
clusterFile = open(clusterFileName, 'w')
for cluster in sortedClusters:
for hit in cluster:
clusterFile.write('%s,' % hit)
clusterFile.write('\n')
clusterFile.close()
'''
#re-create sortedClusters list:
clusterFileName = 'sortedClusters.data'
clusterFile = open(clusterFileName, 'r')
sortedClusters = []
for line in clusterFile:
sortedClusters.append([])
line = line.strip()[0:-1] #take off last comma ;P
for hit in (line.strip().split(',')):
sortedClusters[-1].append(hit)
'''
print 'Store intermediate data:', timer.split()
#output hitsAround file
outputFile = open(conf.conf['hitsPerFrame'], 'w')
frameLength = 200
frameShift = 1
for cluster in sortedClusters:
#grab first and last coordinate from cluster, for each cluster deduce how many theoretical microRNAs were in hitScope
clusterChrom = cluster[0].split(":")[0]
clusterStrand = cluster[0].split(":")[1]
firstCoord = int(cluster[0].split(":")[2])
#print cluster[-1]
lastCoord = int(cluster[-1].split(":")[3])
startCoord = firstCoord
while startCoord < lastCoord:
#count how many hits there are in this range
rangeStart = startCoord - (frameLength/2)
rangeEnd = startCoord + (frameLength/2)
rangeTcc = '%s:%s:%s:%s' % (clusterChrom, clusterStrand, rangeStart, rangeEnd)
overlappedList = compare.compareTwoTcc([rangeTcc], cluster, 2)
hitCount = len(overlappedList)
#output
outputFile.write('%s\t%s\n' % (rangeTcc, hitCount))
startCoord = startCoord + frameShift #check overlap with range
outputFile.close()
print 'Output Hits per Frame:', timer.split()
print 'Overall Time:', timer.report()
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 defineClusters(cName=None):
#Start Timer
timer = cg.cgTimer()
timer.start()
#Get list of mature tccs
conf = cgConfig.getConfig(cName) #passed or default
finalMirFileName = conf.conf['resultsRaw']
matureTccs = compare.tccFileToList(finalMirFileName,
1) # list of all mature micro in tcc
print 'List getting', timer.split()
#make connections dict
matureConnections = compare.makeConnectionsDict(matureTccs)
print 'Make connections:', timer.split()
#Now have to define Clusters...
clusters = []
addedList = []
#I don't think python passes by reference? also I think this function is in the middle because it uses a global variable :P
def createClusters(item=None, mode=None):
if item in addedList:
return 0
elif mode == "top":
clusters.append([item])
addedList.append(
item) ##creates new cluster with the item already stored in it
for connectedItem in matureConnections[item]:
createClusters(connectedItem, "neighbor")
elif mode == "neighbor":
clusters[-1].append(
item) #add this item to the last cluster created
addedList.append(item)
for connectedItem in matureConnections[item]:
createClusters(connectedItem, "neighbor")
for tcc in matureTccs:
createClusters(tcc, "top")
print 'Make Clusters', timer.split()
#Sort Clusters.
sortedClusters = []
for cluster in clusters:
sortedClusters.append(cg.sortTccList(cluster))
print 'Sort Clusters:', timer.split()
#Output sorted cluster file
clusterFileName = conf.conf['sortedClusters']
clusterFile = open(clusterFileName, 'w')
for cluster in sortedClusters:
for hit in cluster:
clusterFile.write('%s,' % hit)
clusterFile.write('\n')
clusterFile.close()
'''
#re-create sortedClusters list:
clusterFileName = 'sortedClusters.data'
clusterFile = open(clusterFileName, 'r')
sortedClusters = []
for line in clusterFile:
sortedClusters.append([])
line = line.strip()[0:-1] #take off last comma ;P
for hit in (line.strip().split(',')):
sortedClusters[-1].append(hit)
'''
print 'Store intermediate data:', timer.split()
#output hitsAround file
outputFile = open(conf.conf['hitsPerFrame'], 'w')
frameLength = 200
frameShift = 1
for cluster in sortedClusters:
#grab first and last coordinate from cluster, for each cluster deduce how many theoretical microRNAs were in hitScope
clusterChrom = cluster[0].split(":")[0]
clusterStrand = cluster[0].split(":")[1]
firstCoord = int(cluster[0].split(":")[2])
#print cluster[-1]
lastCoord = int(cluster[-1].split(":")[3])
startCoord = firstCoord
while startCoord < lastCoord:
#count how many hits there are in this range
rangeStart = startCoord - (frameLength / 2)
rangeEnd = startCoord + (frameLength / 2)
rangeTcc = '%s:%s:%s:%s' % (clusterChrom, clusterStrand,
rangeStart, rangeEnd)
overlappedList = compare.compareTwoTcc([rangeTcc], cluster, 2)
hitCount = len(overlappedList)
#output
outputFile.write('%s\t%s\n' % (rangeTcc, hitCount))
startCoord = startCoord + frameShift #check overlap with range
outputFile.close()
print 'Output Hits per Frame:', timer.split()
print 'Overall Time:', timer.report()
import bioLibCG
import sys
fN = sys.argv[1]
timer = bioLibCG.cgTimer()
timer.start()
loadTime = 0.0
splitTime = 0.0
f = open(fN, 'r')
for line in f:
loadTime += timer.split()
a = line.strip().split('\t')
b = int(a[0])
splitTime += timer.split()
print loadTime
print splitTime
def scanVectorsHist(tccList, cName):
'''Given tcc list --> scan wig files and get histogram values
can be modified to do single/total values...
THIS USES INDEXES!!! = BAD...'''
conf = c.getConfig(cName)
org = conf.conf['organism']
mConf = c.getConfig('Main.conf')
wigDir = mConf.conf['wig%s' % org]
timer = cg.cgTimer()
timer.start()
histDict = {} # tcc: [list values]
for tcc in tccList:
theSplit = ss(tcc, ':')
chrom, strand, tccStart, tccEnd = theSplit[0], theSplit[1],int(theSplit[2]),int(theSplit[3])
#goto correct fild, correct line in index
fN = wigDir + '/Merge.%s.%s.wig.%s.wig' % (org.lower(),strand,chrom)
fNindex = wigDir + '/Merge.%s.%s.wig.%s.wig.index' % (org.lower(),strand,chrom)
#print timer.split()
#get line in index file
iFile = open(fNindex, 'r')
startByte = 'None'
for line in iFile:
beg = int(cg.ss(line)[1])
end = int(cg.ss(line)[2])
if beg <= tccStart < end:
startByte = int(cg.ss(line)[0])
#print 'INDEX', line.strip()
break
iFile.close()
#print timer.split()
#grab value
f = open(fN, 'r')
f.seek(startByte, 0)
stop = False
for line in f:
#print 'Line:', line.strip()
lBeg = int(cg.ss(line)[1])
lEnd = int(cg.ss(line)[2])
lValue = int(cg.ss(line)[3].split('.')[0])
if tccStart > lBeg:
lBeg = tccStart
if tccEnd < lEnd:
lEnd = tccEnd
stop = True
#print timer.split()
for i in range(lBeg, lEnd):
try:
histDict[tcc].append(lValue)
except KeyError: #just for zero...so you don't have to if every time...
histDict[tcc] = [lValue]
if stop: break
f.close()
#print timer.split()
return histDict
