def collectMatureFrames(baseName = None, merLength = None):
#Defaults, Definitions, and Castings
if not baseName:
print 'base name of file family needed!'
return 1
if not merLength:
merLength = 18
merLength = int(merLength)
###############collect IDS that have the same kmer
print 'Collecting kmer IDs'
conf = cgConfig.cgConfig()
idFileName = conf.conf['resultsRaw']
idFile = open(idFileName, 'r')
interFileName = ('./out/%s/' % baseName) + baseName + '.collection.intermediate'
interFile = open(interFileName, 'w')
doneList = []
for line in idFile:
#For each kmer: grab id and Xmer frames --> output
kmerID = line.strip().split('\t')[0].split('.')[0]
if kmerID not in doneList:
doneList.append(kmerID)
matureSeq = line.strip().split('\t')[3]
for frame in returnFrames(matureSeq, merLength):
interFile.write('%s\t%s\n' % (kmerID, frame))
interFile.close()
idFile.close()
def probe(tcc, conf = None):
if not conf:
mConf = c.cgConfig('Main.conf')
smallPath = mConf.conf['smallPath']
chrom, strand, start, end = cg.tccSplit(tcc)
total = 0
for lib in cg.recurseDir(smallPath, end = 'mapped.%s.wig' % strand):
try:
eLevels = stepVectorScan.scanVectorsFile(lib, [tcc])
except:
print lib, 'index failed'
continue
#find highest expression level
highest = 0
for coord in eLevels:
if eLevels[coord] > highest:
highest = eLevels[coord]
if highest > 0:
print lib, highest
total += highest
#print eLevels
print total
def collectMatureFrames(baseName=None, merLength=None):
#Defaults, Definitions, and Castings
if not baseName:
print 'base name of file family needed!'
return 1
if not merLength:
merLength = 18
merLength = int(merLength)
###############collect IDS that have the same kmer
print 'Collecting kmer IDs'
conf = cgConfig.cgConfig()
idFileName = conf.conf['resultsRaw']
idFile = open(idFileName, 'r')
interFileName = ('./out/%s/' %
baseName) + baseName + '.collection.intermediate'
interFile = open(interFileName, 'w')
doneList = []
for line in idFile:
#For each kmer: grab id and Xmer frames --> output
kmerID = line.strip().split('\t')[0].split('.')[0]
if kmerID not in doneList:
doneList.append(kmerID)
matureSeq = line.strip().split('\t')[3]
for frame in returnFrames(matureSeq, merLength):
interFile.write('%s\t%s\n' % (kmerID, frame))
interFile.close()
idFile.close()
def stageTWO(packetNumber=None):
#Caste and defaults
if not packetNumber:
print 'need number of packet to run'
return 1
else:
packetNumber = int(packetNumber)
#####################################################
#Load CONFIGURATION FILE:
#####################################################
conf = c.cgConfig()
print '\nConfiguration:'
for entry in conf.conf:
print '..%s: %s' % (entry, conf.conf[entry])
####################################################
#Files and File Naming
####################################################
#scratchfile
scratchFile = open(
'./scratch.txt', 'w'
) #This is just the file that the warnings for the perl script are redirected to
#directories
outDir = conf.conf['outDirectory']
#Filenaming
pipeName = conf.conf['runName'] + '-' + 's' + conf.conf[
'frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf[
'mirBasePairs'] + '.' + str(packetNumber)
foldOut = outDir + '/' + pipeName + '.folded.frames.txt'
filterOut = outDir + '/' + pipeName + '.filtered.mirs.tsv'
finalOut = outDir + '/' + pipeName + '.FINAL.mirs.tsv'
####################################################
#Pipeline
####################################################
print "\nSTARTING STAGE THREE\n(packet # %s)" % packetNumber
print "-Filtering frames and finding prospective mirs"
subprocess.Popen([
'python', './filter.frames.py', '-i', foldOut, '-g',
conf.conf['genomes'], '-b', conf.conf['mirBasePairs'], '-m',
conf.conf['mirLength'], '-o', filterOut
]).wait()
'''
print "-Running Conservation filter"
subprocess.Popen(['perl', './get_percent_identity_list_fix.pl',
'-g', conf.conf['genomes'],
'-l', filterOut,
'-o', finalOut], stderr = scratchFile).wait()
'''
print "DONE"
def probe(tcc, conf=None):
if not conf:
mConf = c.cgConfig('Main.conf')
smallPath = mConf.conf['smallPath']
chrom, strand, start, end = cg.tccSplit(tcc)
total = 0
for lib in cg.recurseDir(smallPath, end='mapped.%s.wig' % strand):
try:
eLevels = stepVectorScan.scanVectorsFile(lib, [tcc])
except:
print lib, 'index failed'
continue
#find highest expression level
highest = 0
for coord in eLevels:
if eLevels[coord] > highest:
highest = eLevels[coord]
if highest > 0:
print lib, highest
total += highest
#print eLevels
print total
def splitExonsIntrons(cName = None):
mConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
#init
organism = conf.conf['organism']
minOverlap = 50
cHairs = getHairpins.getHairpins() #CID: HAIRPIN
exonList = compare.tccFileToList('%sExons.tcc' % organism, 0)
hairpins = []
for CID in cHairs:
hairpins.append(cHairs[CID])
print 'checking overlaps'
#check which hairpins overlap exons and by how much
exonOverlapped = compare.compareTwoTcc(hairpins, exonList, 1, amount = True)
print ' ', len(exonOverlapped)
print 'removing partial introns'
#remove the ones that didn't overlap more than X:
remList = []
for tcc, oAmount in exonOverlapped:
if oAmount < minOverlap:
remList.append([tcc, oAmount])
for item in remList:
exonOverlapped.remove(item)
print ' ', len(exonOverlapped), 'out of', len(cHairs.keys())
#get CIDs of exons
exonCIDs = []
for tcc, oAmount in exonOverlapped:
for CID in cHairs:
if cHairs[CID] == tcc:
exonCIDs.append(str(CID))
#Open sorted predictions and write lines with CIDs to respective files
predFile = open(conf.conf['resultsSorted'], 'r')
exonFile = open(conf.conf['resultsSorted'] + '.exons', 'w')
intronFile = open(conf.conf['resultsSorted'] + '.introns', 'w')
for line in predFile:
if line.split('\t')[7] in exonCIDs:
exonFile.write(line)
else:
intronFile.write(line)
predFile.close()
exonFile.close()
intronFile.close()
def splitExonsIntrons(cName=None):
mConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
#init
organism = conf.conf['organism']
minOverlap = 50
cHairs = getHairpins.getHairpins() #CID: HAIRPIN
exonList = compare.tccFileToList('%sExons.tcc' % organism, 0)
hairpins = []
for CID in cHairs:
hairpins.append(cHairs[CID])
print 'checking overlaps'
#check which hairpins overlap exons and by how much
exonOverlapped = compare.compareTwoTcc(hairpins, exonList, 1, amount=True)
print ' ', len(exonOverlapped)
print 'removing partial introns'
#remove the ones that didn't overlap more than X:
remList = []
for tcc, oAmount in exonOverlapped:
if oAmount < minOverlap:
remList.append([tcc, oAmount])
for item in remList:
exonOverlapped.remove(item)
print ' ', len(exonOverlapped), 'out of', len(cHairs.keys())
#get CIDs of exons
exonCIDs = []
for tcc, oAmount in exonOverlapped:
for CID in cHairs:
if cHairs[CID] == tcc:
exonCIDs.append(str(CID))
#Open sorted predictions and write lines with CIDs to respective files
predFile = open(conf.conf['resultsSorted'], 'r')
exonFile = open(conf.conf['resultsSorted'] + '.exons', 'w')
intronFile = open(conf.conf['resultsSorted'] + '.introns', 'w')
for line in predFile:
if line.split('\t')[7] in exonCIDs:
exonFile.write(line)
else:
intronFile.write(line)
predFile.close()
exonFile.close()
intronFile.close()
def createMultiTrackDir(dirName, organism):
'''THIS DIFFERS FROM ABOVE BECAUSE IT DOESN't REQUIRE META INFO
IT JUST MAKES A MERGED WIG FOR EVERYTHING IN THE DIRECTORY'''
mainConf = c.cgConfig('Main.conf')
fileList = []
for file in cg.recurseDir(dirName, end = '.mapped'):
fileList.append(file)
#make merged wig
if organism == 'human':
chroms = cg.humanChromosomes
assembly = 'hg19'
elif organism == 'mouse':
chroms = cg.mouseChromosomes
assembly = 'mm9'
elif organism == 'zebrafish':
chroms = cg.zebrafishChromosomes
assembly = 'danRer6'
print 'Making Bed File vectors'
cvg = HTSeq.GenomicArray(chroms, stranded=True, typecode='i')
for fName in fileList:
alignment_file = HTSeq.BowtieReader(fName)
for alngt in alignment_file:
if alngt.aligned:
cvg.add_value( 1, alngt.iv ) #iv is the genomic interval..
bedNamePos = dirName + '/Merge.' + organism + '.1.wig'
bedNameNeg = dirName + '/Merge.' + organism + '.-1.wig'
print 'Writing Bed File'
cvg.write_bedgraph_file(bedNamePos, "+" )
cvg.write_bedgraph_file(bedNameNeg, "-" )
#Now extend it
updateWigLength(bedNamePos, assembly)
updateWigLength(bedNameNeg, assembly)
#Now Sort it.
cgSort.wigSort(bedNamePos)
cgSort.wigSort(bedNameNeg)
def finish(packetNumber): #init conf = c.cgConfig() #directories outDir = conf.conf['outDirectory'] #Filenaming pipeName = conf.conf['runName'] + '-' + 's' + conf.conf['frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf['mirBasePairs'] + '.' + str(packetNumber) extractOut = outDir + '/' + pipeName + '.folding.frames.tsv' splitDirectory = outDir + '/' + pipeName + '/' foldOut = outDir + '/' + pipeName + '.folded.frames.txt' print '-Stitching %s files back into one (%s)' % (conf.conf['numSplitFiles'], packetNumber) subprocess.Popen(['python', './stitchdb.py', '-b', pipeName, '-n', conf.conf['numSplitFiles'], '-o', outDir + '/', '-i', splitDirectory]).wait()
def queryNumJobsQ(user):
#init
mainConf = cgConfig.cgConfig('Main.conf')
scratchFileName = mainConf.conf['qstatScratch']
#output qstat info to scratch file
sFile = open(scratchFileName, 'w')
subprocess.Popen(['qstat', '-u', user], stdout=sFile).wait()
sFile.close()
#parse scratchfile, check how many times usr name appears
sFile = open(scratchFileName, 'r')
userCount = 0
for line in sFile:
if user in line:
userCount = userCount + 1
sFile.close()
return userCount
def queryNumJobsQ(user):
#init
mainConf = cgConfig.cgConfig('Main.conf')
scratchFileName = mainConf.conf['qstatScratch']
#output qstat info to scratch file
sFile = open(scratchFileName, 'w')
subprocess.Popen(['qstat', '-u', user], stdout=sFile).wait()
sFile.close()
#parse scratchfile, check how many times usr name appears
sFile = open(scratchFileName, 'r')
userCount = 0
for line in sFile:
if user in line:
userCount = userCount + 1
sFile.close()
return userCount
def finish(packetNumber):
#init
conf = c.cgConfig()
#directories
outDir = conf.conf['outDirectory']
#Filenaming
pipeName = conf.conf['runName'] + '-' + 's' + conf.conf[
'frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf[
'mirBasePairs'] + '.' + str(packetNumber)
extractOut = outDir + '/' + pipeName + '.folding.frames.tsv'
splitDirectory = outDir + '/' + pipeName + '/'
foldOut = outDir + '/' + pipeName + '.folded.frames.txt'
print '-Stitching %s files back into one (%s)' % (
conf.conf['numSplitFiles'], packetNumber)
subprocess.Popen([
'python', './stitchdb.py', '-b', pipeName, '-n',
conf.conf['numSplitFiles'], '-o', outDir + '/', '-i', splitDirectory
]).wait()
def createTrackInDir(dirName):
'''Every Q function has a corresponding shell script
Make wig file for all mapped files, for all organisms'''
wrapperShell = '/home/chrisgre/scripts/mapping/createTrack.sh'
mainConf = c.cgConfig('Main.conf')
metaFileName = mainConf.conf['metaFileName']
for file in cg.recurseDir(dirName, end = '.mapped'):
#check if mouse or human
baseFName = cg.getBaseFileName(file)
baseFName = baseFName.split('.')[0]
metaDict = cg.getMetaFileDict(metaFileName)
org = 'None'
if baseFName in metaDict:
if metaDict[baseFName][1] == 'NONE':
print ' NO ORG KNOWN FOR', file
continue
else:
org = metaDict[baseFName][1]
print ' USING ORG', org, file
#check if there is an organism, must check due to files not in metaFile
if org == 'None':
print ' NO org (not in meta file)', file
continue
while True:
#submit job if there are less than ten
if clusterCheck.queryNumJobsQ('chrisgre') < 1000:
#subprocess.Popen(['qsub', '-q', 'xiao', '-l', 'mem=4G', '-V', '-o', mainConf.conf['outLog'], '-e', mainConf.conf['errorLog'], wrapperShell, file, org ])
subprocess.Popen(['qsub', '-V', '-o', mainConf.conf['outLog'], '-e', mainConf.conf['errorLog'], wrapperShell, file, org ])
#time.sleep(.5) #give it time to update qstat
break
else:#wait 10 secs...
time.sleep(20)
def stageTWO(packetNumber=None):
#Caste and defaults
if not packetNumber:
print 'need number of packet to run'
return 1
else:
packetNumber = int(packetNumber)
#####################################################
#Load CONFIGURATION FILE:
#####################################################
conf = c.cgConfig()
'''
print '\nConfiguration:'
for entry in conf.conf:
print '..%s: %s' % (entry, conf.conf.conf[entry])
'''
####################################################
#Files and File Naming
####################################################
#scratchfile
scratchFile = open(
'./scratch.txt', 'w'
) #This is just the file that the warnings for the perl script are redirected to
#directories
outDir = conf.conf['outDirectory']
#Filenaming
#!!!THIS IS DIFFERENENT THAN MIRPIPEPARA!!! (pipename is the name of packet...)
pipeName = conf.conf['runName'] + '-' + 's' + conf.conf[
'frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf[
'mirBasePairs'] + '.' + str(packetNumber)
extractOut = outDir + '/' + pipeName + '.folding.frames.tsv'
splitDirectory = outDir + '/' + pipeName + '/'
foldOut = outDir + '/' + pipeName + '.folded.frames.txt'
####################################################
#Pipeline
####################################################
print "\nSTARTING STAGE TWO\n(packet # %s)" % packetNumber
if int(conf.conf['numSplitFiles']) == 1: #if you want to run on one node
print '-Folding Frames using RNAfold on SINGLE NODE'
subprocess.Popen(['./RNAfold.sh', extractOut, foldOut]).wait()
else: #Else parallize
print '-Splitting folding frames into %s files (%s)' % (
conf.conf['numSplitFiles'], packetNumber)
subprocess.Popen([
'python', './splitdb.py', '-i', extractOut, '-b', pipeName, '-n',
conf.conf['numSplitFiles'], '-d', splitDirectory
]).wait()
print '-Submitting %s seperate jobs to cluster (%s)' % (
conf.conf['numSplitFiles'], packetNumber)
subprocess.Popen([
'python', './parFold.py', '-b', pipeName, '-n',
conf.conf['numSplitFiles'], '-d', splitDirectory
]).wait()
#Get Job ID's:
parseFile = open('%s.jobsinfo.txt' % pipeName, 'r') #!!! edit this?
jobIDs = []
for line in parseFile:
jobIDs.append(line.split(' ')[2])
#check if right number were submitted
if len(jobIDs) == int(conf.conf['numSplitFiles']):
print '....jobs were submitted correctly'
print 'DONE (%s)' % packetNumber
# get the hairpin values
# can also reuse elements from this script to get CID elements...
import bioLibCG as cg
from bioLibCG import ss
import cgConfig as c
mConf = c.cgConfig("Main.conf")
conf = c.cgConfig()
def getHairpins(fN):
predFile = open(fN, "r")
# populate CID:hairpin range
cHairs = {}
for line in predFile:
# get cluster ID
CID = ss(line)[7]
hairpin = ss(line)[2]
if CID in cHairs:
# check if the starts and ends need to be stretched
hStart = int(ss(cHairs[CID], ":")[2])
hEnd = int(ss(cHairs[CID], ":")[3])
start = int(ss(hairpin, ":")[2])
end = int(ss(hairpin, ":")[3])
if start < hStart:
hStart = start
if end > hEnd:
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 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 exonNoisy(cName = None):
#init
mConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
cHairs = getHairpins.getHairpins(conf.conf['resultsExons']) #CID: HAIRPIN
organism = conf.conf['organism']
geneSetFolder = mConf.conf['geneSets%s' % organism]
#make prediction overlap hitmap
print 'Making prediction list'
predList = []
for CID in cHairs:
hPin = cHairs[CID]
predList.append(hPin)
if compare.checkIfOverlaps(predList):
predList = compare.collapseOverlaps(predList)
#make genes for Ensemble/make list of tccs for exons.
print 'Creating gene set'
ensGenes = cgGenes.createGeneSetFromFile(geneSetFolder + '/ensemblAllExons.tsv')
print ' loaded # genes:', len(ensGenes.set)
#collect levels for each haipin region
print '[Checking all levels]'
cidLevels = {}
for CID in cHairs:
print CID
hPin = cHairs[CID]
#for each hairpin, --> find overlapping transcripts in same gene
overlappingGenes = ensGenes.geneOverlaps([hPin])
if len(overlappingGenes) > 0:
gIDs = [gene.id for gene in overlappingGenes]
allTccs = ensGenes.getTccsFromGIDs(gIDs)
if compare.checkIfOverlaps:
print ' Overlaps...collapsing'
allTccs = compare.collapseOverlaps(allTccs)
else:
print 'NO GENE OVERLAPS!!!!!', CID, hPin
#filter out my predictions.
print ' Filtering out predictions'
checkList = compare.subtractTwoTccLists(allTccs, predList)
#Get Expression level for gene.
print ' Retrieving Expression levels:', cg.getTccListTotalLength(checkList)
levels = []
hPinLevels = stepVectorScan.scanVectorsHist(checkList, cName)
for hPin in hPinLevels:
levels.extend(hPinLevels[hPin])
cidLevels[CID] = levels
#output levels to file
print 'Outputting to file'
#find longest
longest = 0
for CID in cidLevels:
length = len(cidLevels[CID])
if length > longest:
longest = length
sortedKeys = cidLevels.keys()
sortedKeys.sort()
#print sortedKeys
newLines = []
for j in range(0, longest): #how many lines are there
newLine = []
for CID in sortedKeys:
if len(cidLevels[CID]) > j:# add it
newLine.append(str(cidLevels[CID][j]))
else:
newLine.append('NA')
newLines.append('\t'.join(newLine) + '\n')
outFileN = conf.conf['exonNoiseData']
outFile = open(outFileN, 'w')
outFile.write('\t'.join(sortedKeys) + '\n')
outFile.writelines(newLines)
outFile.close()
def intronNoisy(cName = None):
mConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
#init
cHairs = getHairpins.getHairpins(conf.conf['resultsIntrons']) #CID: HAIRPIN
organism = conf.conf['organism']
exonList = compare.tccFileToList('%sExons.tcc' % organism, 0)
slide = 1000
#make prediction overlap hitmap
predMap = {}
predList = []
for CID in cHairs:
hPin = cHairs[CID]
predList.append(hPin)
#collapse Overlaps
print ' collapsing predictions'
predList = compare.collapseOverlaps(predList)
print ' collapsing exons'
exonList = compare.collapseOverlaps(exonList)
#collect levels for each hairpin region
cidLevels = {}
for CID in cHairs:
print CID
hPin = cHairs[CID]
chrom = ss(hPin, ':')[0]
strand = ss(hPin, ':')[1]
start = int(ss(hPin, ':')[2])
end = int(ss(hPin, ':')[3])
scanStart = start - slide
scanEnd = end + slide
scanRange = []
scanRange.append('%s:%s:%s:%s' % (chrom, strand, scanStart, start))
scanRange.append('%s:%s:%s:%s' % (chrom, strand, end, scanEnd))
print scanRange
scanRange = compare.subtractTwoTccLists(scanRange, predList)
scanRange = compare.subtractTwoTccLists(scanRange, exonList)
levels = []
print ' Retrieving Expression levels:', cg.getTccListTotalLength(scanRange)
levels = []
hPinLevels = stepVectorScan.scanVectorsHist(scanRange, cName)
for hPin in hPinLevels:
levels.extend(hPinLevels[hPin])
cidLevels[CID] = levels
#output levels to file
#find longest
longest = 0
for CID in cidLevels:
length = len(cidLevels[CID])
if length > longest:
longest = length
sortedKeys = cidLevels.keys()
sortedKeys.sort()
newLines = []
for j in range(0, longest): #how many lines are there
newLine = []
for CID in sortedKeys:
if len(cidLevels[CID]) > j:# add it
newLine.append(str(cidLevels[CID][j]))
else:
newLine.append('NA')
newLines.append('\t'.join(newLine) + '\n')
outFileN = conf.conf['intronNoiseData']
outFile = open(outFileN, 'w')
outFile.write('\t'.join(sortedKeys) + '\n')
outFile.writelines(newLines)
outFile.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 intronNoisy(cName=None):
mConf = c.cgConfig('Main.conf')
conf = c.getConfig(cName)
#init
cHairs = getHairpins.getHairpins(
conf.conf['resultsIntrons']) #CID: HAIRPIN
organism = conf.conf['organism']
exonList = compare.tccFileToList('%sExons.tcc' % organism, 0)
slide = 1000
#make prediction overlap hitmap
predMap = {}
predList = []
for CID in cHairs:
hPin = cHairs[CID]
predList.append(hPin)
#collapse Overlaps
print ' collapsing predictions'
predList = compare.collapseOverlaps(predList)
print ' collapsing exons'
exonList = compare.collapseOverlaps(exonList)
#collect levels for each hairpin region
cidLevels = {}
for CID in cHairs:
print CID
hPin = cHairs[CID]
chrom = ss(hPin, ':')[0]
strand = ss(hPin, ':')[1]
start = int(ss(hPin, ':')[2])
end = int(ss(hPin, ':')[3])
scanStart = start - slide
scanEnd = end + slide
scanRange = []
scanRange.append('%s:%s:%s:%s' % (chrom, strand, scanStart, start))
scanRange.append('%s:%s:%s:%s' % (chrom, strand, end, scanEnd))
print scanRange
scanRange = compare.subtractTwoTccLists(scanRange, predList)
scanRange = compare.subtractTwoTccLists(scanRange, exonList)
levels = []
print ' Retrieving Expression levels:', cg.getTccListTotalLength(
scanRange)
levels = []
hPinLevels = stepVectorScan.scanVectorsHist(scanRange, cName)
for hPin in hPinLevels:
levels.extend(hPinLevels[hPin])
cidLevels[CID] = levels
#output levels to file
#find longest
longest = 0
for CID in cidLevels:
length = len(cidLevels[CID])
if length > longest:
longest = length
sortedKeys = cidLevels.keys()
sortedKeys.sort()
newLines = []
for j in range(0, longest): #how many lines are there
newLine = []
for CID in sortedKeys:
if len(cidLevels[CID]) > j: # add it
newLine.append(str(cidLevels[CID][j]))
else:
newLine.append('NA')
newLines.append('\t'.join(newLine) + '\n')
outFileN = conf.conf['intronNoiseData']
outFile = open(outFileN, 'w')
outFile.write('\t'.join(sortedKeys) + '\n')
outFile.writelines(newLines)
outFile.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 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 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()
import fastQTypes, cgConfig
import bioLibCG as cg
import subprocess
import os, time
import clusterCheck
#init
mainConf = cgConfig.cgConfig('Main.conf')
metaFileNames = [mainConf.conf['metaFileName']]
wrapperShell = '/home/chrisgre/scripts/smallRNAProcessing/clipAdapter.sh'
def clipAdapter(fName, adapter = None, validate = False, oName = None, overwrite = True):
#Check to see if the file exists:
putativeN = fName.replace('.fastq','.clipped.fastq')
if os.path.isfile(putativeN):
if overwrite:
print ' Overwriting file', putativeN
os.remove(putativeN)
else:
print ' \nNOT OVERWRITING FILE', putativeN
return 1
#If the adapter is none, try to find it in the small.meta file
if adapter is None:
baseFName = cg.getBaseFileName(fName) + '.counts'
for metaFileName in metaFileNames:
mFile = open(metaFileName, 'r')
for line in mFile:
fields = line.strip().split('\t')
if baseFName == fields[0]:
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()
###This is the MAIN script to compare small RNA libraries to the mature microRNA sequences. It first calls
##collectMatureFrames and then calls compareFinished to do the actual comparisons.
from bioLibCG import *
import os
from collectMF import *
from compareCounts import *
import cgConfig
#init
conf = cgConfig.cgConfig()
baseName = conf.conf['baseName']
merLength = conf.conf['merLength']
inDirectory = conf.conf['inDirectory']
smallPath = conf.conf['smallPath']
mainConf = cgConfig.cgConfig('Main.conf')
smallPath = mainConf.conf['smallPath']
merLength = int(merLength)
####Make directory for files to go into
filePath = 'out/%s' % baseName
if baseName in os.listdir('out/'):
#Delete all contents
for file in os.listdir(filePath):
os.remove('%s/%s' % (filePath, file))
else:
os.mkdir(filePath)
###Get ids and Xmer frames
collectMatureFrames(baseName, merLength)
def stageTWO(packetNumber = None):
#Caste and defaults
if not packetNumber:
print 'need number of packet to run'
return 1
else:
packetNumber = int(packetNumber)
#####################################################
#Load CONFIGURATION FILE:
#####################################################
conf = c.cgConfig()
print '\nConfiguration:'
for entry in conf.conf:
print '..%s: %s' % (entry, conf.conf[entry])
####################################################
#Files and File Naming
####################################################
#scratchfile
scratchFile = open('./scratch.txt', 'w') #This is just the file that the warnings for the perl script are redirected to
#directories
outDir = conf.conf['outDirectory']
#Filenaming
pipeName = conf.conf['runName'] + '-' + 's' + conf.conf['frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf['mirBasePairs'] + '.' + str(packetNumber)
foldOut = outDir + '/' + pipeName + '.folded.frames.txt'
filterOut = outDir + '/' + pipeName + '.filtered.mirs.tsv'
finalOut = outDir + '/' + pipeName + '.FINAL.mirs.tsv'
####################################################
#Pipeline
####################################################
print "\nSTARTING STAGE THREE\n(packet # %s)" % packetNumber
print "-Filtering frames and finding prospective mirs"
subprocess.Popen(['python', './filter.frames.py',
'-i', foldOut,
'-g', conf.conf['genomes'],
'-b', conf.conf['mirBasePairs'],
'-m', conf.conf['mirLength'],
'-o', filterOut]).wait()
'''
print "-Running Conservation filter"
subprocess.Popen(['perl', './get_percent_identity_list_fix.pl',
'-g', conf.conf['genomes'],
'-l', filterOut,
'-o', finalOut], stderr = scratchFile).wait()
'''
print "DONE"
#generate number of reads per prediction.
import bioLibCG as cg
import cgConfig
conf = cgConfig.cgConfig()
resultsFile = open(conf.conf['resultsSorted'], 'r')
resultsFile = open(conf.conf['onePerLine'], 'r')
#make header for R
print 'density'
for line in resultsFile:
print line.strip().split('\t')[12]
#mirPipe using parrallel cpu's
import subprocess
import time, os
import cgConfig as c
startTime = time.time()
#####################################################
#Load CONFIGURATION FILE:
#####################################################
conf = c.cgConfig()
print '\nConfiguration:'
for entry in conf.conf:
print '..%s: %s' % (entry, conf.conf[entry])
####################################################
#Files and File Naming
####################################################
#scratchfile
scratchFile = open('./scratch.txt', 'w') #This is just the file that the warnings for the perl script are redirected to
#directories
outDir = conf.conf['outDirectory']
#Filenaming
pipeName = conf.conf['runName'] + '-' + 's' + conf.conf['frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf['mirBasePairs']
conservedOut = outDir + '/' + pipeName + '.ALL.conserved.kmers.tsv'
import cgConfig as c
import bioLibCG as cg
import cgSort
mConf = c.cgConfig('Main.conf')
smallPath = mConf.conf['smallPath']
smallPath = '/home/chrisgre/smallLibs/WIGS/zebrafish'
#grab everything - NOT WIG MERGES...
smallLibs = cg.recurseDir(smallPath, end = '.wig')
smallLibs.extend(cg.recurseDir(smallPath, end = '.wig'))
for lib in smallLibs:
print 'sorting', lib
cgSort.wigSort(lib)
#mirPipe using parrallel cpu's
import subprocess
import time, os
import cgConfig as c
startTime = time.time()
#####################################################
#Load CONFIGURATION FILE:
#####################################################
conf = c.cgConfig()
print '\nConfiguration:'
for entry in conf.conf:
print '..%s: %s' % (entry, conf.conf[entry])
####################################################
#Files and File Naming
####################################################
#scratchfile
scratchFile = open(
'./scratch.txt', 'w'
) #This is just the file that the warnings for the perl script are redirected to
#directories
outDir = conf.conf['outDirectory']
#Filenaming
pipeName = conf.conf['runName'] + '-' + 's' + conf.conf[
'frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf[
def stageTWO(packetNumber = None):
#Caste and defaults
if not packetNumber:
print 'need number of packet to run'
return 1
else:
packetNumber = int(packetNumber)
#####################################################
#Load CONFIGURATION FILE:
#####################################################
conf = c.cgConfig()
'''
print '\nConfiguration:'
for entry in conf.conf:
print '..%s: %s' % (entry, conf.conf.conf[entry])
'''
####################################################
#Files and File Naming
####################################################
#scratchfile
scratchFile = open('./scratch.txt', 'w') #This is just the file that the warnings for the perl script are redirected to
#directories
outDir = conf.conf['outDirectory']
#Filenaming
#!!!THIS IS DIFFERENENT THAN MIRPIPEPARA!!! (pipename is the name of packet...)
pipeName = conf.conf['runName'] + '-' + 's' + conf.conf['frameStep'] + 'k' + conf.conf['kmerLength'] + 'b' + conf.conf['mirBasePairs'] + '.' + str(packetNumber)
extractOut = outDir + '/' + pipeName + '.folding.frames.tsv'
splitDirectory = outDir + '/' + pipeName + '/'
foldOut = outDir + '/' + pipeName + '.folded.frames.txt'
####################################################
#Pipeline
####################################################
print "\nSTARTING STAGE TWO\n(packet # %s)" % packetNumber
if int(conf.conf['numSplitFiles']) == 1: #if you want to run on one node
print '-Folding Frames using RNAfold on SINGLE NODE'
subprocess.Popen(['./RNAfold.sh', extractOut, foldOut]).wait()
else: #Else parallize
print '-Splitting folding frames into %s files (%s)' % (conf.conf['numSplitFiles'], packetNumber)
subprocess.Popen(['python', './splitdb.py',
'-i', extractOut,
'-b', pipeName,
'-n', conf.conf['numSplitFiles'],
'-d', splitDirectory]).wait()
print '-Submitting %s seperate jobs to cluster (%s)' % (conf.conf['numSplitFiles'], packetNumber)
subprocess.Popen(['python', './parFold.py',
'-b', pipeName,
'-n', conf.conf['numSplitFiles'],
'-d', splitDirectory]).wait()
#Get Job ID's:
parseFile = open('%s.jobsinfo.txt' % pipeName, 'r') #!!! edit this?
jobIDs = []
for line in parseFile:
jobIDs.append(line.split(' ')[2])
#check if right number were submitted
if len(jobIDs) == int(conf.conf['numSplitFiles']):
print '....jobs were submitted correctly'
print 'DONE (%s)' % packetNumber
def createMultiTrack(dirName, organism):
'''merge all mapped tracks in directory and create a single wig file'''
mainConf = c.cgConfig('Main.conf')
metaFileName = mainConf.conf['metaFileName']
fileList = []
for file in cg.recurseDir(dirName, end = '.mapped'):
#check if mouse or human SHOULD PUT INTO A STD FUNCTION FOR META FILE
#check if mouse or human
baseFName = cg.getBaseFileName(file, naked= True)
metaDict = cg.getMetaFileDict(metaFileName)
org = 'None'
if baseFName in metaDict:
if metaDict[baseFName][1] == 'NONE':
print ' NO ORG KNOWN FOR', file
continue
elif not metaDict[baseFName][1] == organism:
print ' NOT ORGANISM RUNNING', file
continue
else:
org = metaDict[baseFName][1]
print ' USING ORG', org, file
#check if there is an organism, must check due to files not in metaFile
if org == 'None':
print ' NO org (not in meta file)', file
continue
#only make wig file for organism asked for
if not org == organism:
continue
#if it is right organism and has mapped file then add
fileList.append(file)
#make merged wig
if organism == 'human':
chroms = cg.humanChromosomes
assembly = 'hg19'
elif organism == 'mouse':
chroms = cg.mouseChromosomes
assembly = 'mm9'
elif organism == 'zebrafish':
chroms = cg.zebrafishChromosomes
assembly = 'danRer6'
print 'Making Bed File vectors'
cvg = HTSeq.GenomicArray(chroms, stranded=True, typecode='i')
for fName in fileList:
alignment_file = HTSeq.BowtieReader(fName)
for alngt in alignment_file:
if alngt.aligned:
cvg.add_value( 1, alngt.iv ) #iv is the genomic interval..
bedNamePos = dirName + '/Merge.' + organism + '.1.wig'
bedNameNeg = dirName + '/Merge.' + organism + '.-1.wig'
print 'Writing Bed File'
cvg.write_bedgraph_file(bedNamePos, "+" )
cvg.write_bedgraph_file(bedNameNeg, "-" )
#Now extend it
updateWigLength(bedNamePos, assembly)
updateWigLength(bedNameNeg, assembly)
#Now Sort it.
cgSort.wigSort(bedNamePos)
cgSort.wigSort(bedNameNeg)
