def updateAvgNumSS(oFN):
bn = os.path.basename(oFN)
print 'basename', bn
oID_numSS = {}
numSims = 100
print 'getting avg for %s simulations' % numSims
for i in range(0, numSims):
#simFN = '/home/chrisgre/scripts/simulations/simsk50FilteredMasked/simulation.%s/%s' % (i, bn)
#simFN = '/home/chrisgre/scripts/simulations/simsk50Fix/simulation.%s/%s' % (i, bn)
#simFN = '/home/chrisgre/scripts/simulations/mm9/simulation.%s/%s' % (i, bn)
#simFN = '/home/chrisgre/scripts/simulations/hg19.hela/simulation.%s/%s' % (i, bn)
simFN = '/home/chrisgre/scripts/simulations/hg19.U87/simulation.%s/%s' % (
i, bn)
osNX = cgNexusFlat.Nexus(simFN, cgOriginRNAFlat.OriginRNA)
osNX.load(['numSignificantSequences'])
for oID in osNX.numSignificantSequences:
oID_numSS[oID] = oID_numSS.get(
oID, 0) + osNX.numSignificantSequences[oID]
#now save it
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['avgNumSS'])
for oID in oNX.avgNumSS:
totalNum = oID_numSS.get(oID, 0)
avgNum = float(totalNum) / float(numSims)
oNX.avgNumSS[oID] = avgNum
oNX.save()
def updateTargetIDsFiltered(oFN, aFN, rn=None, tn=None):
'''CAUTION: NO SELECTION BEING MADE!!!'''
#load the data
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['filteredTargets'], [rn, tn])
'''
#get ids of alignments I need (set)
oIDs = set()
for oID in oNX.filteredTargets: oIDs.add(oID)
'''
#load only alignments I need
'''c = {'sID' : lambda x: x in oIDs}'''
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['sID'])
#clear targets that are there.
for oID in oNX.filteredTargets:
oNX.filteredTargets[oID] = []
#update the targets for oRNAs
for aID in aNX.sID:
oID = aNX.sID[aID]
try:
oNX.filteredTargets[oID].append(aID)
except KeyError: #another process is taking care of this one
pass
#save
oNX.save()
def appendTInfoFlat(aFN, dFN, rn = None, tn = None):
aNX = cgNexusFlat.Nexus(aFN, cgAlignment)
aNX.load(['tID', 'tTcc', 'transcriptOverlap', 'tELevel', 'context', 'repeat', 'targetSequence', 'gScore'], [rn, tn])
dNX = cgNexusFlat.Nexus(dFN, cgDegPeak.Peak)
dNX.load(['tOverlap', 'eLevel', 'tcc', 'context', 'repeatStatus', 'sequence', 'gScore'])
tID_aIDs = {}
for aID in aNX.tID:
tID_aIDs.setdefault(aNX.tID[aID], []).append(aID)
for dID in dNX.tcc:
for aID in tID_aIDs.get(dID, list()):
aNX.tTcc[aID] = dNX.tcc[dID]
aNX.tELevel[aID] = dNX.eLevel[dID]
aNX.transcriptOverlap[aID] = dNX.tOverlap[dID]
aNX.context[aID] = dNX.context[dID]
aNX.repeat[aID] = dNX.repeatStatus[dID]
aNX.gScore[aID] = dNX.gScore[dID]
aNX.targetSequence[aID] = dNX.sequence[dID]
#aNX.repeatCount[aID] = dNX.repeatCount[dID]
#aNX.totalContig[aID] = dNX.totalContig[dID]
aNX.save()
def updateAvgNumTargets(oFN):
bn = os.path.basename(oFN)
print 'basename', bn
oID_numTargets = {}
for i in range(0, 10):
#simFN = '/home/chrisgre/scripts/simulations/simsk50FilteredMasked/simulation.%s/%s' % (i, bn)
simFN = '/home/chrisgre/scripts/simulations/simsk50/simulation.%s/%s' % (
i, bn)
print simFN
osNX = cgNexusFlat.Nexus(simFN, cgOriginRNAFlat.OriginRNA)
osNX.load(['filteredTargets'])
for oID in osNX.filteredTargets:
currTargets = oID_numTargets.get(oID, 0)
oID_numTargets[oID] = currTargets + len(osNX.filteredTargets[oID])
#now save it
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['avgNumSimulationTargets'])
for oID in oNX.avgNumSimulationTargets:
totalNum = oID_numTargets.get(oID, 0)
avgNum = float(totalNum) / float(10.0)
oNX.avgNumSimulationTargets[oID] = avgNum
oNX.save()
def eLevelHistogram(oFN, aFN, oRNA=True):
oRNA = 'True' in oRNA
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['filteredTargets', 'eLevel'])
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['tELevel'])
histValues = []
for oID in oNX.eLevel:
if oRNA:
histValues.append(oNX.eLevel[oID])
else:
for aID in oNX.filteredTargets[oID]:
histValues.append(aNX.tELevel[aID])
histVals = [math.log(x, 10) for x in histValues]
plt.hist(histVals, 50)
type = 'oRNA'
if not oRNA: type = 'Targets (degradome)'
plt.title('Expression Level for %s' % type)
plt.xlabel('log(Expression Level)')
plt.ylabel('Number of %s' % type)
plt.show()
def countRepeatStatusTargets(oFN, aFN, oContext=None, oType=None):
if oContext == 'None': oContext = None
if oType == 'None': oType = None
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(
['snrSS', 'context', 'transcriptType', 'filteredTargets', 'gScore'],
[rn, tn])
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['context', 'gScore', 'repeat'], [rn, tn])
context_rStatuss = {}
for oID in oNX.context:
if oNX.snrSS[oID] < 2.00: continue
#gather targets' context info if oRNA is okay
for aID in oNX.filteredTargets[oID]:
aCon = aNX.context[aID]
rStatus = aNX.repeat[aID]
context_rStatuss.setdefault(aCon, []).append(rStatus)
#plot
for context in context_rStatuss:
print context, context_rStatuss[context].count(
True), context_rStatuss[context].count(False)
def gZipContextECDF(oFN, aFN, imgName, oContext=None, oType=None):
if oContext == 'None': oContext = None
if oType == 'None': oType = None
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(
['snrSS', 'context', 'transcriptType', 'filteredTargets', 'gScore'],
[rn, tn])
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['context', 'gScore'], [rn, tn])
context_gzips = {}
for oID in oNX.context:
if oNX.snrSS[oID] < 2.00: continue
#gather targets' context info if oRNA is okay
for aID in oNX.filteredTargets[oID]:
aCon = aNX.context[aID]
gScore = aNX.gScore[aID]
context_gzips.setdefault(aCon, []).append(gScore)
#plot
for context in context_gzips:
plt.hist(context_gzips[context],
bins=10000,
cumulative=True,
histtype='step',
normed=True,
label='%s' % context)
plt.legend()
plt.savefig(imgName, bbox_inches='tight', pad_inches=1)
def updateTargetIDs(oFN, aFN, rn=None, tn=None):
#load the data
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['targets'], [rn, tn])
#get ids of alignments I need (set)
oIDs = set()
for oID in oNX.targets:
oIDs.add(oID)
#load only alignments I need
c = {'sID': lambda x: x in oIDs}
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['sID'], conditions=c)
#clear targets that are there.
for oID in oNX.targets:
oNX.targets[oID] = []
#update the targets for oRNAs
for aID in aNX.sID:
oID = aNX.sID[aID]
oNX.targets[oID].append(aID)
#save
oNX.save()
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 targetContextPercentageVsExpression(oFN,
aFN,
oContext=None,
oType=None,
rn=None,
tn=None):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['context', 'transcriptType', 'filteredTargets'])
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['context', 'tELevel'])
context_level_count = {}
for lev in range(50, 500, 50):
for oID in oNX.context:
#filter oID types here
con = oNX.context[oID]
typ = oNX.transcriptType[oID]
if oContext:
if oContext != con: continue
if oType:
if oType != typ: continue
#gather targets' context info if oRNA is okay
for aID in oNX.filteredTargets[oID]:
eLevel = aNX.tELevel[aID]
if eLevel < lev: continue
aCon = aNX.context[aID]
context_level_count[aCon][
lev] = context_level_count.setdefault(aCon, {}).get(
lev, 0) + 1
#fracs = pieFractions(counts)
plots_labels = [[], []]
for con in context_level_count:
x = []
y = []
sortedLevs = sorted(context_level_count[con].keys())
for lev in sortedLevs:
x.append(lev)
y.append(context_level_count[con][lev])
plots_labels[0].append(plt.plot(
x,
y,
))
plots_labels[1].append(con)
#plot
plt.legend(plots_labels[0], plots_labels[1])
plt.title(
'oRNA Targets\' Context Proportion Stability w/ Expression Increase')
plt.xlabel('Degradome Expression Cutoff')
plt.ylabel('Number of oRNA Targets')
plt.show()
def phastScoreByNT(oFN, aFN, oIDFilter=None):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['phastScores', 'snrSS', 'filteredTargets'])
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['mismatchPositions'])
misPositionsX = []
misScoresY = []
positionsX = []
scoresY = []
for oID in oNX.phastScores:
avgScore = sum(oNX.phastScores[oID]) / float(len(oNX.phastScores[oID]))
#filter
if (avgScore < .90) or (oNX.snrSS[oID] < 2):
continue
if oIDFilter:
if oID != int(oIDFilter):
continue
misPositions = set()
#get consolidated mismatches
for aID in oNX.filteredTargets[oID]:
for mPos in aNX.mismatchPositions[aID]:
misPositions.add(mPos)
for i, pScore in enumerate(oNX.phastScores[oID]):
if i in misPositions:
misPositionsX.append(
i + 1.2
) #1 is for 0BASE, .2 is for differentiating between mis and reg
misScoresY.append(pScore)
else:
positionsX.append(i + 1)
scoresY.append(pScore)
highestNT = max(positionsX)
for i in range(1, highestNT + 1):
plt.axvspan(i - .15, i + .35, facecolor='g', alpha=.25)
plt.plot(positionsX, scoresY, 'bo')
plt.plot(misPositionsX, misScoresY, 'ro')
plt.ylim(0, 1.1)
plt.xlim(0, 24)
plt.title('Conservation by Position of Conserved oRNA')
plt.ylabel('PhastCons Score')
plt.xlabel('Nucleotide Position')
plt.show()
def filterTargets(oFN,
aFN,
inTranscript,
misLevel,
centerLevel,
minCenterLevel,
rn=None,
tn=None):
if inTranscript == 'True': inTranscript = True
if inTranscript == 'False': inTranscript = False
misLevel, centerLevel, minCenterLevel = int(misLevel), int(
centerLevel), float(minCenterLevel)
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['filteredTargets', 'targets'], [rn, tn])
print inTranscript, misLevel, centerLevel, minCenterLevel, oFN, aFN, rn, tn
#make selection set
targets = set()
for oID in oNX.targets:
for target in oNX.targets[oID]:
targets.add(target)
c = {'ID': lambda x: x in targets}
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['transcriptOverlap', 'mismatchStatus', 'centerExpression'],
conditions=c)
for oID in oNX.filteredTargets:
oNX.filteredTargets[oID] = []
for aID in oNX.targets[oID]:
#transcriptOverlap
if inTranscript:
if not aNX.transcriptOverlap[aID]:
#print 'tOverlap Fail', cgAlignment.pretty#print(alignment)
continue
#misLevel
if aNX.mismatchStatus[aID][misLevel]:
#print 'mismatch Fail', cgAlignment.pretty#print(alignment)
continue
#centerLevel
if aNX.centerExpression[aID][centerLevel] < minCenterLevel:
#print 'expression Fail', cgAlignment.pretty#print(alignment)
continue
oNX.filteredTargets[oID].append(aID)
oNX.save()
def conservedHisto(oFN):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['phastScores'])
scores = []
for oID in oNX.phastScores:
avgScore = sum(oNX.phastScores[oID]) / float(len(oNX.phastScores[oID]))
#scores.extend(oNX.phastScores[oID])
scores.append(avgScore)
print len(scores)
plt.title('PhastCons Scores By Nucleotide')
plt.title('PhastCons Scores By oRNA')
plt.ylabel('Number of Nucleotides')
plt.ylabel('Number of oRNA')
plt.xlabel('PhastCons Score')
plt.xlabel('PhastCons Score Average')
plt.hist(scores, 50)
plt.show()
def updateTranscriptOverlap(oFN, wigDir, chrom, strand, rn=None, tn=None):
oNX = cgNexusFlat.Nexus(oFN, cgDegPeak.Peak)
oNX.load(['tOverlap', 'tcc'], [rn, tn])
#load the AS wig file for this degradome strand
if strand == '1':
strand = '-1'
else:
strand = '1'
coord_transcripts = cgWig.loadSingleWigTranscript(wigDir, chrom, strand,
'transcript')
for oID in oNX.tOverlap:
tChrom, tStrand, start, end = bioLibCG.tccSplit(oNX.tcc[oID])
if tStrand == '1':
tStrand = '-1'
else:
tStrand = '1'
if tChrom != chrom or tStrand != strand: continue
oNX.tOverlap[oID] = False
for i in xrange(start, end + 1):
if i in coord_transcripts:
oNX.tOverlap[oID] = True
break
oNX.save()
def markMismatchedPairs(aFN, rn=None, tn=None):
#make mismatchDict
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['mismatchStatus', 'mismatchPositions'], [rn, tn])
lowRange = range(
8, 12) # remember the small locations are 0-based, so 10 is 9
midRange = range(7, 13)
highRange = range(6, 14)
for aID in aNX.mismatchStatus:
aNX.mismatchStatus[aID] = [False, False, False]
#check mismatches
for i in lowRange:
if i in aNX.mismatchPositions[aID]:
aNX.mismatchStatus[aID][0] = True
break
for i in midRange:
if i in aNX.mismatchPositions[aID]:
aNX.mismatchStatus[aID][1] = True
break
for i in highRange:
if i in aNX.mismatchPositions[aID]:
aNX.mismatchStatus[aID][2] = True
break
aNX.save()
def getGeneGC(genePropFN, rn=None, tn=None):
myGF = gf.GenomeFetch('hg19')
NX = cgNexusFlat.Nexus(genePropFN, geneProperty)
NX.load([
'geneName', 'geneChrom', 'geneStrand', 'geneStarts', 'geneEnds',
'geneGCContent'
], [rn, tn])
for id in NX.ids:
spanPairs = zip(NX.geneStarts[id], NX.geneEnds[id])
spanTccs = [
'%s:%s:%s:%s' %
(NX.geneChrom[id], NX.geneStrand[id], pair[0], pair[1])
for pair in spanPairs
]
totalSequenceLength = 0
numGC = 0
for tcc in spanTccs:
seq = myGF.getSequence(tcc)
totalSequenceLength += len(seq)
numGC += seq.count('G') + seq.count('C')
GCC = float(numGC) / totalSequenceLength
NX.geneGCContent[id] = GCC
NX.save()
def updateGeneData(geneRanges, genePropFN):
myGF = gf.GenomeFetch('hg19')
NX = cgNexusFlat.Nexus(genePropFN, geneProperty)
NX.load(['geneName', 'geneChrom', 'geneStrand', 'geneStarts', 'geneEnds'])
#make inverse dictionary
gName_nID = {}
for id in NX.ids:
gName_nID[NX.geneName[id]] = id
f = open(geneRanges, 'r')
for line in f:
ls = line.strip().split('\t')
sChrom, sStrand = ls[1], ls[2]
geneName = ls[0]
geneStarts = [int(x) for x in ls[3].split(',')]
geneEnds = [int(x) for x in ls[4].split(',')]
#get id
nID = gName_nID.get(geneName, None)
if nID:
NX.geneChrom[nID] = sChrom
NX.geneStrand[nID] = sStrand
NX.geneStarts[nID] = geneStarts
NX.geneEnds[nID] = geneEnds
f.close()
NX.save()
def filterOrigin(oFN, rn=None, tn=None):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load([
'filteredTargets', 'endContigLength', 'totalContigLength',
'sequenceDuplicate', 'passedFilter', 'entropy'
], [rn, tn])
for oID in oNX.entropy:
oNX.passedFilter[oID] = False
if oNX.entropy[oID] < 1.15:
continue
if oNX.endContigLength[oID] > 6:
continue
if oNX.totalContigLength[oID] > 6:
continue
if oNX.sequenceDuplicate[oID]:
continue
if not oNX.filteredTargets[oID]:
continue
oNX.passedFilter[oID] = True
oNX.save()
def getGeneLength(geneRanges, genePropFN):
NX = cgNexusFlat.Nexus(genePropFN, geneProperty)
NX.load(['geneName', 'geneLength'])
#make inverse dictionary
gName_nID = {}
for id in NX.ids:
gName_nID[NX.geneName[id]] = id
f = open(geneRanges, 'r')
for line in f:
ls = line.strip().split('\t')
sChrom, sStrand = ls[1], ls[2]
geneName = ls[0]
geneStarts = [int(x) for x in ls[3].split(',')]
geneEnds = [int(x) for x in ls[4].split(',')]
spanPairs = zip(geneStarts, geneEnds)
totalLength = sum([pair[1] - pair[0] for pair in spanPairs])
#get id
nID = gName_nID.get(geneName, None)
if nID:
NX.geneLength[nID] = totalLength
f.close()
NX.save()
def totalSNRSS(oFN, SNRToggle=False):
if SNRToggle == 'True':
SNRToggle = True
else:
SNRToggle = False
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['snrSS', 'numSignificantSequences', 'avgNumSS'])
highSNRs = []
for oID in oNX.snrSS:
snrSS = oNX.snrSS[oID]
if snrSS > 2:
highSNRs.append(snrSS)
if SNRToggle:
try:
avgSNR = sum(highSNRs) / len(highSNRs)
print avgSNR,
except:
print '0.0',
else:
print len(highSNRs),
def totalSNR(oFN):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(
['snr', 'filteredTargets', 'avgNumSimulationTargets', 'passedFilter'])
totalRun = 0
totalSim = 0
simsTotals = []
n = 0
highSNR = 0
for oID in oNX.avgNumSimulationTargets:
#filter out
if not oNX.passedFilter[oID]:
continue
#collect stats
totalRun += len(oNX.filteredTargets[oID])
simsTotal = oNX.avgNumSimulationTargets[oID] * 10
simsTotals.append(simsTotal)
totalSim += oNX.avgNumSimulationTargets[oID]
if oNX.snr[oID] > 2:
highSNR += 1
n += 1
print oFN
print 'Total Number Targets for my run:', totalRun
print 'Total Number Targets for Simulations:', totalSim
print 'SNR', float(totalRun) / float(totalSim)
print 'Total oRNA:', n, 'Total oRNA w/ SNR > 2:', highSNR
print '\n'
def updateFiltered(oFN):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load([
'filteredTargets', 'endContigLength', 'totalContigLength', 'entropy',
'sequenceDuplicate', 'passedFilter'
])
for oID in oNX.passedFilter:
oNX.passedFilter[oID] = False
if len(oNX.filteredTargets[oID]) == 0:
continue
if oNX.endContigLength[oID] > 6:
continue
if oNX.totalContigLength[oID] > 6:
continue
if oNX.entropy[oID] < 1.2:
continue
if oNX.sequenceDuplicate[oID]:
continue
#if it passed, update
oNX.passedFilter[oID] = True
oNX.save()
def correlationSNR(oFN):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['phastScores', 'snrSS'])
snrX = []
scoreY = []
for oID in oNX.phastScores:
snr = oNX.snrSS[oID]
avgScore = sum(oNX.phastScores[oID]) / float(len(oNX.phastScores[oID]))
snrX.append(snr)
scoreY.append(avgScore)
'''
for pScore in oNX.phastScores[oID]:
snrX.append(snr)
scoreY.append(pScore)
'''
conserved = [True if x > .8 else False for x in scoreY]
print conserved.count(True)
print len(conserved)
plt.title('SNR vs PhastCons Score')
plt.ylabel('Avg PhastCons Score of oRNA')
plt.xlabel('SNR')
plt.plot(snrX, scoreY, 'ro')
plt.show()
def oRNAContextPie(oFN, imgName):
'''REMEMBER!!! Have to do with grouped results...'''
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['context', 'snrSS'])
context_count = {}
for oID in oNX.context:
if oNX.snrSS[oID] < 2.00: continue
con = oNX.context[oID]
print con
context_count[con] = context_count.get(con, 0) + 1
labels = sorted(context_count.keys())
counts = [context_count[x] for x in labels]
fracs = pieFractions(counts)
#add numbers to labels
labels = ['%s (%s)' % (x, context_count[x]) for x in labels]
#plot
plt.title('Context of oRNA (results > 2.00 SNR)')
plt.pie(fracs, labels=labels, shadow=True)
plt.savefig(imgName, bbox_inches='tight', pad_inches=1)
def updateEndContig(oFN, rn = None, tn = None):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['sequence', 'endContigLength'], [rn, tn])
for oID in oNX.sequence:
seq = oNX.sequence[oID]
#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
oNX.endContigLength[oID] = highest
oNX.save()
def loadSeqs(seqFN):
seqNX = cgNexusFlat.Nexus(seqFN, Seq)
seqNX.load(['length', 'sequence'])
print seqNX.length[100000], seqNX.sequence[100000]
print 'done loading'
def markMismatchedPairs(aFN, rn=None, tn=None):
#make mismatchDict
aNX = cgNexusFlat.Nexus(aFN, cgAlignmentFlat.cgAlignment)
aNX.load(['mismatchStatus', 'mismatchPositions'], [rn, tn])
for aID in aNX.mismatchStatus:
aNX.mismatchStatus[aID] = [False, False, False]
lowRange = range(9, 13)
midRange = range(8, 14)
highRange = range(7, 15)
#check mismatches
for i in lowRange:
if i in aNX.mismatchPositions[aID]:
aNX.mismatchStatus[aID][0] = True
break
for i in midRange:
if i in aNX.mismatchPositions[aID]:
aNX.mismatchStatus[aID][1] = True
break
for i in highRange:
if i in aNX.mismatchPositions:
aNX.mismatchStatus[aID][2] = True
break
aNX.save()
def updateSharedTargets(oFN, rn=None, tn=None):
'''Just because there are duplicate sequences does not mean that
the genomic position of each results is the correct one. The
targets for each genomic position should be the same as the targets
for each duplicate sequence
make set of targets for each oID --> set each oid's targets'''
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['sequence', 'filteredTargets'], [rn, tn])
knownSeq_targets = {}
#create oID groups and target sets.
for oID in oNX.sequence:
currSeq = oNX.sequence[oID]
#add targets to set
for tID in oNX.filteredTargets[oID]:
knownSeq_targets.setdefault(currSeq, set()).add(tID)
for oID in oNX.sequence:
currSeq = oNX.sequence[oID]
newTargets = list(knownSeq_targets.get(currSeq, set()))
oNX.filteredTargets[oID] = newTargets
oNX.save()
def oRNATypePie(oFN, imgName):
oNX = cgNexusFlat.Nexus(oFN, cgOriginRNAFlat.OriginRNA)
oNX.load(['snrSS', 'transcriptType', 'transcriptTypes'], [rn, tn])
context_count = {}
for oID in oNX.transcriptType:
#con = oNX.transcriptType[oID]
if oNX.snrSS[oID] < 2.00: continue
cons = oNX.transcriptTypes[oID]
for con in cons:
context_count[con] = context_count.get(con, 0) + 1
#context_count[con] = context_count.get(con, 0) + 1
labels = sorted(context_count.keys())
counts = [context_count[x] for x in labels]
fracs = pieFractions(counts)
#add numbers to labels
labels = ['%s (%s)' % (x, context_count[x]) for x in labels]
#plot
plt.pie(fracs, labels=labels, shadow=True)
plt.savefig(imgName, bbox_inches='tight', pad_inches=1)
def updateContext(oFN, wigDir, chrom, strand, rn=None, tn=None):
oNX = cgNexusFlat.Nexus(oFN, degPeak.degPeak)
oNX.load(['context', 'tcc'], [rn, tn])
print 'loading wig'
coord_contexts = cgWig.loadSingleWigContext(wigDir, chrom, strand,
'context')
print 'done loading'
ds = bioLibCG.dominantSpotter([
'C_EXON', 'C_3UTR', 'C_5UTR', 'NC_EXON', 'NC_3UTR', 'NC_5UTR',
'C_INTRON', 'NC_INTRON', 'INTER'
])
for oID in oNX.tcc:
oChrom, oStrand, start, end = bioLibCG.tccSplit(oNX.tcc[oID])
#deg wigs is AS to actual clipping site
if oStrand == '1':
oStrand = '-1'
else:
oStrand = '1'
if oChrom == chrom and oStrand == strand:
contexts = coord_contexts.get(start, 'INTER').split(',')
oNX.context[oID] = ds.spotItem(contexts)
oNX.save()