def test(gFN):
map = cgSeqMod.loadCodonMap("hg19")
geneSet = cgGenes3.createGeneSetEditing(gFN)
for transcript in geneSet.transcripts:
if "_coding" in transcript.tType:
try:
print transcript.id
print transcript.getMRNA(coding=True)
print cgSeqMod.translateRNA(transcript.getMRNA(coding=True), map)
except:
print "fail"
def test(gFN):
map = cgSeqMod.loadCodonMap('hg19')
geneSet = cgGenes3.createGeneSetEditing(gFN)
for transcript in geneSet.transcripts:
if '_coding' in transcript.tType:
try:
print transcript.id
print transcript.getMRNA(coding=True)
print cgSeqMod.translateRNA(transcript.getMRNA(coding=True),
map)
except:
print 'fail'
def testit(gFN):
geneSet = cgGenes3.createGeneSetEditing(gFN)
map = cgSeqMod.loadCodonMap('hg19')
for gene in geneSet.genes:
for transcript in gene.transcripts:
try:
print ''
mRNA = transcript.getMRNA(coding = True)
i = transcript.getRelativePositionMRNA(35872409)
if i == -1:
continue
print transcript.id
print i
print mRNA[:i], mRNA[i], mRNA[i + 1:]
print cgSeqMod.translateRNA(mRNA, map)
except:
pass
def updateLocationBasedTargets(editFN, contextFN, miLocationFN, gFN):
eSites = cgEdit.loadEditingSites(editFN)
cgEdit.updateContextEditingSites(eSites, contextFN) #puts the UTR, EXON in eSite.context
geneSet = cgGenes3.createGeneSetEditing(gFN)
tName_t = {}
for t in geneSet.transcripts:
tName_t[t.id] = t
tName_miInfo = {}
f = open(miLocationFN, 'r')
for line in f:
ls = line.strip().split('\t')
tName = ls[0]
miName = ls[1]
loc = int(ls[2])
tName_miInfo.setdefault(tName, []).append([miName, loc])
for eSite in eSites:
if '3UTR' not in eSite.context:
continue
for tName in eSite.transcripts:
if tName in tName_miInfo:
t = tName_t[tName]
for info in tName_miInfo[tName]:
miName = info[0]
loc = info[1]
#get the position of e site in mrna for this transcript
ePosition = t.getRelativePositionMRNA(eSite.coordinate, coding = False)
print tName, miName, loc, ePosition
if loc - 22 <= ePosition <= loc:
print tName, miName, '%s:%s' % (eSite.chromosome, eSite.coordinate)
pass
def getCoords(gFN):
geneSet = cgGenes3.createGeneSetEditing(gFN)
for gName in geneSet.set:
gene = geneSet.set[gName]
# find the longest utr...
longestUTR = None
longestT = None
longest = 0
for transcript in gene.transcripts:
l = 0
if transcript.utr3 is None:
continue
for utrPair in transcript.utr3:
l += utrPair[1] - utrPair[0] + 1
if l > longest:
longestUTR = transcript.utr3
longestT = transcript
# Get the coordinate ends/etc
starts, ends = [], []
if longestUTR is None:
continue
for utrPair in longestUTR:
starts.append(utrPair[0])
ends.append(utrPair[1])
starts.sort()
ends.sort()
startS = ",".join([str(x) for x in starts])
endS = ",".join([str(x) for x in ends])
print "%s\t%s\t%s\t%s\t%s\t%s" % (
transcript.id,
transcript.parent,
transcript.chromosome,
transcript.strand,
startS,
endS,
)
def checkIfSeedPresent(gName_micros, gFN):
gf = GenomeFetch.GenomeFetch('hg19')
print 'loading gene set'
geneSet = cgGenes3.createGeneSetEditing(gFN)
print '....done loading'
outF = open('utrSeeds', 'w')
done = {}
for gName in gName_micros:
micros = gName_micros[gName]
for transcript in geneSet.set[gName].transcripts:
utrCoords = []
if len(transcript.utr3) == 0:
continue
for utrPair in transcript.utr3:
utrCoords.extend(utrPair)
utrCoords.sort()
start, end = utrCoords[0], utrCoords[1]
chrom = transcript.chromosome
strand = transcript.strand
checkSeq = gf.get_seq_from_to(chrom, start, end, strand)
print ''
print transcript.id
print checkSeq
checkSeq = checkSeq.replace('T', 'U')
for micro in micros:
findings = checkSeq.find(micro.comSeed)
if findings != -1:
outF.write('%s\t%s\t%s\n' % ( transcript.id, micro.name, transcript.parent))
found = findings
def getCoords(gFN):
geneSet = cgGenes3.createGeneSetEditing(gFN)
for gName in geneSet.set:
gene = geneSet.set[gName]
#find the longest utr...
longestUTR = None
longestT = None
longest = 0
for transcript in gene.transcripts:
l = 0
if transcript.utr3 is None: continue
for utrPair in transcript.utr3:
l += utrPair[1] - utrPair[0] + 1
if l > longest:
longestUTR = transcript.utr3
longestT = transcript
#Get the coordinate ends/etc
starts, ends = [], []
if longestUTR is None: continue
for utrPair in longestUTR:
starts.append(utrPair[0])
ends.append(utrPair[1])
starts.sort()
ends.sort()
startS = ','.join([str(x) for x in starts])
endS = ','.join([str(x) for x in ends])
print '%s\t%s\t%s\t%s\t%s\t%s' % (transcript.id, transcript.parent,
transcript.chromosome,
transcript.strand, startS, endS)
def newExpression(eFN, gFN, contextFN, targetFN, rpkmFN):
"""docstring for newExpression"""
#print 'loading editing/genes'
eSites = cgEdit.loadEditingSites(eFN)
geneSet = cgGenes3.createGeneSetEditing(gFN)
#print 'making joints'
#joints
eID_eSite = {}
for eSite in eSites:
eID_eSite[eSite.ID] = eSite
gID_eIDs = {}
eID_tTypes = {}
f = open(contextFN, 'r')
for line in f:
ls = line.strip().split('\t')
gID = ls[1]
eID = int(ls[0])
tType = ls[3]
if gID in gID_eIDs:
if eID not in gID_eIDs[gID]:
gID_eIDs[gID].append(eID)
else:
gID_eIDs[gID] = [eID]
if eID in eID_tTypes:
eID_tTypes[eID].append(tType)
else:
eID_tTypes[eID] = [tType]
#print 'updating target sites'
#update targetting for eSites:
f = open(targetFN, 'r')
for line in f:
ls = line.strip().split('\t')
eID = int(ls[0])
before = ls[3].split(',')
if before == ['None']: before = []
after = ls[4].split(',')
if after == ['None']: after = []
eID_eSite[eID].before = before
eID_eSite[eID].after = after
#scrolldict
gene_eSites = {}
for gID in gID_eIDs:
try:
gene = geneSet.set[gID]
except KeyError:
#print gID, 'not in geneSet'
pass
gene_eSites[gene] = []
#get eSites
for eID in gID_eIDs[gID]:
eSite = eID_eSite[eID]
gene_eSites[gene].append(eSite)
#print 'updating gene target site info'
#update before/after editing target sites for GENES
createdGenes = []
destroyedGenes = []
histoVals = []
for gene in gene_eSites:
gene.before = []
gene.after = []
for eSite in gene_eSites[gene]:
if '3UTR' in eID_tTypes[eSite.ID]:
for micro in eSite.before:
gene.before.append('%s %s:%s' % (micro, eSite.ID, eSite.coordinate))
for micro in eSite.after:
gene.after.append('%s %s:%s' % (micro, eSite.ID, eSite.coordinate))
if len(gene.before) > len(gene.after):
destroyedGenes.append(gene)
if len(gene.after) > len(gene.before):
createdGenes.append(gene)
change = len(gene.after) - len(gene.before)
if (len(gene.after) != 0) or (len(gene.before) != 0):
histoVals.append(change)
#print the created/destroyed sites
for gene in createdGenes:
print gene.before
print gene.after
for gene in destroyedGenes:
print gene.before
print gene.after
'''
plt.title('Target Site Changes Due To Editing')
plt.xlabel('Change in Number of Target Sites After Editing')
plt.ylabel('Number of Genes')
plt.hist(histoVals, 6)
plt.show()
return 0
'''
#print out microRNA gene List
uniqueMicros = {}
for gene in createdGenes:
for micro in gene.after:
uniqueMicros[micro] = 1
for gene in destroyedGenes:
for micro in gene.before:
uniqueMicros[micro] = 1
for micro in uniqueMicros:
#print micro
pass
#print 'loading rpkm'
gName_ratio = getERatioDict(rpkmFN)
eChanges = []
for gene in createdGenes:
try:
ratio = gName_ratio[gene.id]
except KeyError:
#print gene.id, 'not in RPKM file --> not expressed'
pass
eChange = math.log(ratio, 2)
eChanges.append(eChange)
eChanges2 = []
for gene in destroyedGenes:
try:
ratio = gName_ratio[gene.id]
except KeyError:
#print gene.id, 'not in RPKM file --> not expressed'
pass
eChange = math.log(ratio, 2)
eChanges2.append(eChange)
#Now plot the histogram
plt.hist(eChanges, 40, cumulative = True, histtype = 'step', normed = True, label = 'Created')
plt.hist(eChanges2, 40, cumulative = True, histtype = 'step', normed = True, label = 'Destroyed')
plt.legend()
plt.title('eCDF of Genes with Target Site Changes in 3UTRs')
plt.xlabel('log2(RPKM KD/ RPKM CONTROL)')
plt.ylabel('Fraction of Genes')
plt.show()
def updateContext(editFN, geneSetFN, outFN, refBase='A'):
print refBase
#Load Transcripts and Editing Sites
print 'loading editing sites'
eSites = cgEdit.loadEditingSites(editFN, refBase)
print 'loading gene set'
geneSet = cgGenes3.createGeneSetEditing(geneSetFN)
#make the eSites 0 based
for eSite in eSites:
#redo coordinate and tcc
eSite.coordinate = eSite.coordinate - 1
eSite.tcc = bioLibCG.makeTcc(eSite.chromosome, eSite.strand,
eSite.coordinate, eSite.coordinate)
#Create Joint dictionaries
print 'creating joint dictionaries'
eJoint = {} #tcc : eSite
for eSite in eSites:
eJoint[eSite.tcc] = eSite
tJoint = {} # tcc : [transcript, ...]
for transcript in geneSet.transcripts:
if transcript.tcc in tJoint:
tJoint[transcript.tcc].append(transcript)
else:
tJoint[transcript.tcc] = [transcript]
#Overlap tccs
print 'overlapping joints'
##make new 0-based keys
tccOverlaps = compareData.getIndividualOverlaps(eJoint.keys(),
tJoint.keys(), 1)
print 'creating final dictionary'
#create final dictionary containing {edit sites : [transcript, ..]}
eSiteTranscripts = {} # edit site: [transcript, ..]
for eTcc in tccOverlaps:
eSite = eJoint[eTcc]
eSiteTranscripts[eSite] = []
for tTcc in tccOverlaps[eTcc]:
eSiteTranscripts[eSite].extend(tJoint[tTcc])
print 'get context info'
#Go through each site and find out what it overlaps, and if it is in a coding region...
fOut = open(outFN, 'w')
for eSite in eSiteTranscripts:
if len(eSiteTranscripts[eSite]) == 0:
#label intergenic
tType = 'INTER'
codingFlag = 'NC'
fOut.write('%s\t%s\t%s\t%s\t%s\t%s\n' %
(eSite.ID, 'NONE', 'NONE', tType, codingFlag, 'NONE'))
continue
for transcript in eSiteTranscripts[eSite]:
codingTranscript = '_coding' in transcript.tType
tType = None
codingFlag = None
tTypes = [
x[1] for x in transcript.getOverlappingElements(eSite.tcc)
]
if '3UTR' in tTypes:
tType = '3UTR'
elif '5UTR' in tTypes:
tType = '5UTR'
else:
tType = tTypes[0] #has to be one thing...exon or intron
#This only works because UTR takes precedence over EXON in TYPE.
if tType == 'EXON':
if codingTranscript:
codingFlag = 'C'
else:
codingFlag = 'NC'
else:
codingFlag = 'NC'
fOut.write('%s\t%s\t%s\t%s\t%s\t%s\n' %
(eSite.ID, transcript.parent, transcript.id, tType,
codingFlag, transcript.tType))
#fOut.write('%s:%s:%s\t%s\t%s\t%s\t%s\n' % (eSite.chromosome, eSite.strand, eSite.coordinate, transcript.parent, transcript.id, tType, codingFlag))
fOut.close()
def doit(fN):
geneSet = cgGenes3.createGeneSetEditing(fN)
for transcript in geneSet.transcripts:
if transcript.id == 'NM_031422':
dumpObj.dumpObj(transcript)
def makeContextPieBetter(contextFN, gFN, eFN, passedInfo, outFN):
if passedInfo[0] == passedInfo[1]:
print passedInfo, 'no such thing'
return 0
print 'loading geneSet'
geneSet = cgGenes3.createGeneSetEditing(gFN)
typeCount = {'EXON': 0, '3UTR': 0, '5UTR': 0, 'INTRON': 0, 'NONG': 0, 'NONT': 0}
#joint
tID_transcript = {}
for transcript in geneSet.transcripts:
tID_transcript[transcript.id] = transcript
eID_Info = {}
f = open(eFN, 'r')
for line in f:
ls = line.strip().split('\t')
eID = int(ls[13])
chrom = ls[0]
coord = ls[1]
newCoord = '%s:%s' % (chrom, coord)
nEdited = ls[4]
nTotal = ls[5]
eID_Info[eID] = [newCoord, nEdited, nTotal]
eID_gName = {}
eID_tTypes = {}
f = open(contextFN, 'r')
for line in f:
ls = line.strip().split('\t')
eID = int(ls[0])
gName = ls[1]
eID_gName[eID] = gName
tType = ls[3]
tName = ls[2]
if tType == 'INTER':
continue
transcript = tID_transcript[tName]
tInfo = [tType, transcript]
if eID in eID_tTypes:
eID_tTypes[eID].append(tInfo)
else:
eID_tTypes[eID] = [tInfo]
eID_finalType = {}
for eID in eID_tTypes:
highestType = None
for tInfo in eID_tTypes[eID]:
tType = tInfo[0]
transcript = tInfo[1]
tCoding = True
if '_coding' not in transcript.tType:
tCoding = False
gCoding = True
if '_coding' not in transcript.gType:
gCoding = False
#print transcript.id, tType, tCoding, gCoding
if tType == 'EXON':
if tCoding:
highestType = 'EXON'
break
else:
if gCoding:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR']:
highestType = 'NONT'
else:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR', 'NONT']:
highestType = 'NONG'
elif tType == '3UTR':
if tCoding:
if gCoding:
highestType = '3UTR'
else:
highestType = '3UTR'
else:
if gCoding:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR']:
highestType = 'NONT'
else:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR', 'NONT']:
highestType = 'NONG'
elif tType == '5UTR':
if tCoding:
if gCoding:
if highestType not in ['3UTR']:
highestType = '5UTR'
else:
if highestType not in ['3UTR']:
highestType = '5UTR'
else:
if gCoding:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR']:
highestType = 'NONT'
else:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR', 'NONT']:
highestType = 'NONG'
elif tType == 'INTRON':
if tCoding:
if gCoding:
if highestType not in ['3UTR', '5UTR']:
highestType = 'INTRON'
else:
if highestType not in ['3UTR', '5UTR']:
highestType = 'INTRON'
else:
if gCoding:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR']:
highestType = 'NONT'
else:
if highestType not in ['EXON', 'INTRON', '3UTR', '5UTR', 'NONT']:
highestType = 'NONG'
eID_finalType[eID] = highestType
typeCount[highestType] += 1
outF = open(outFN, 'w')
for eID in eID_finalType:
outF.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\n' % (eID_Info[eID][0].split(':')[0], eID_Info[eID][0].split(':')[1], passedInfo, eID_gName[eID], eID_finalType[eID], eID_Info[eID][1], eID_Info[eID][2]))
return 0
outF = open(outFN, 'w')
for type in typeCount:
outF.write('%s\t%s\n' % (type, typeCount[type]))
return 0
#get fractions of each type
types = ['EXON', '3UTR', '5UTR', 'INTRON', 'NONG', 'NONT']
fracs = [typeCount['EXON'], typeCount['3UTR'], typeCount['5UTR'], typeCount['INTRON'], typeCount['NONG'], typeCount['NONT']]
#print fracs
labels = ['Exons (%s)' % fracs[0], '3\'UTR (%s)' % fracs[1], '5\'UTR (%s)' % fracs[2], 'Introns (%s)' % fracs[3], 'Noncoding Gene (%s)' % fracs[4], 'Noncoding Transcript (%s)' % fracs[5]]
theSum = fracs[0] + fracs[1] + fracs[2] + fracs[3] + fracs[4] + fracs[5]
fracs = [float(x)/theSum for x in fracs]
#print fracs
explode=(0.1, 0.1, 0.1, 0.1, .1, .1)
pie(fracs, explode=explode, labels=labels, autopct='%1.1f%%', shadow=True)
title('Editing Site Genomic Location', bbox={'facecolor':'1.0', 'pad':10})
show()
def makeTargetExpressionHistogram(eFN, targetFN, contextFN, geneFN, eChangeFN):
print 'loading expression ratios'
gName_eChange = getERatioDict(eChangeFN)
print 'loading eSites and Transcripts'
eSites = cgEdit.loadEditingSites(eFN)
geneSet = cgGenes3.createGeneSetEditing(geneFN)
print 'making joint dicts and loading extra data'
#joint
eID_eSite = {}
for eSite in eSites:
eID_eSite[eSite.ID] = eSite
#joint
tID_gName = {}
for transcript in geneSet.transcripts:
tID_gName[transcript.id] = transcript.parent
#load context data
f = open(contextFN, 'r')
eID_tID = {} # eID: tID
tID_tType = {}
for line in f:
ls = line.strip().split('\t')
eID = int(ls[0])
tID = ls[1]
tType = ls[2]
tID_tType[tID] = tType
if eID in eID_tID:
eID_tID[eID].append(tID)
else:
eID_tID[eID] = [tID]
f.close()
print 'analyzing'
#Get created or destroyed
f = open(targetFN, 'r')
altered = []
for line in f:
ls = line.strip().split('\t')
#created/destroyed
if (ls[3] != 'None') and (ls[4] == 'None'):
altered.append(int(ls[0]))
f.close()
print 'number of created/destroyed sites:', len(altered)
alteredSites = []
for id in altered:
alteredSites.append(eID_eSite[id])
eChanges = []
gDone = []
#Get gene names for each eSite
for eSite in alteredSites:
genes = []
for tID in eID_tID[eSite.ID]:
if tID == 'NONE':
continue
gName = tID_gName[tID]
if tID_tType[tID] != '3UTR':
print 'Not 3UTR', tID
continue
if gName not in genes:
genes.append(gName)
if len(genes) > 1:
print 'more than one gene for eSite...', genes
continue
if gName in gDone:
continue
else:
gDone.append(gName)
#Now add expression to HistoGram List...
if gName in gName_eChange:
eChange = gName_eChange[gName]
else:
print 'gene not in expression list', gName
continue
eChange = gName_eChange[gName]
eChange = math.log(eChange, 2)
eChanges.append(eChange)
#Now plot the histogram
plt.hist(eChanges, 40)
plt.xlabel('log2(RPKM KD/ RPKM CONTROL)')
plt.ylabel('# Genes')
plt.show()
def updateContext(editFN, geneSetFN, outFN, refBase = 'A'):
print refBase
#Load Transcripts and Editing Sites
print 'loading editing sites'
eSites = cgEdit.loadEditingSites(editFN, refBase)
print 'loading gene set'
geneSet = cgGenes3.createGeneSetEditing(geneSetFN)
#make the eSites 0 based
for eSite in eSites:
#redo coordinate and tcc
eSite.coordinate = eSite.coordinate - 1
eSite.tcc = bioLibCG.makeTcc(eSite.chromosome, eSite.strand, eSite.coordinate, eSite.coordinate)
#Create Joint dictionaries
print 'creating joint dictionaries'
eJoint = {} #tcc : eSite
for eSite in eSites:
eJoint[eSite.tcc] = eSite
tJoint = {} # tcc : [transcript, ...]
for transcript in geneSet.transcripts:
if transcript.tcc in tJoint:
tJoint[transcript.tcc].append(transcript)
else:
tJoint[transcript.tcc] = [transcript]
#Overlap tccs
print 'overlapping joints'
##make new 0-based keys
tccOverlaps = compareData.getIndividualOverlaps(eJoint.keys(), tJoint.keys(), 1)
print 'creating final dictionary'
#create final dictionary containing {edit sites : [transcript, ..]}
eSiteTranscripts = {} # edit site: [transcript, ..]
for eTcc in tccOverlaps:
eSite = eJoint[eTcc]
eSiteTranscripts[eSite] = []
for tTcc in tccOverlaps[eTcc]:
eSiteTranscripts[eSite].extend(tJoint[tTcc])
print 'get context info'
#Go through each site and find out what it overlaps, and if it is in a coding region...
fOut = open(outFN, 'w')
for eSite in eSiteTranscripts:
if len(eSiteTranscripts[eSite]) == 0:
#label intergenic
tType = 'INTER'
codingFlag = 'NC'
fOut.write('%s\t%s\t%s\t%s\t%s\t%s\n' % (eSite.ID, 'NONE', 'NONE', tType, codingFlag, 'NONE'))
continue
for transcript in eSiteTranscripts[eSite]:
codingTranscript = '_coding' in transcript.tType
tType = None
codingFlag = None
tTypes = [ x[1] for x in transcript.getOverlappingElements(eSite.tcc)]
if '3UTR' in tTypes:
tType = '3UTR'
elif '5UTR' in tTypes:
tType = '5UTR'
else:
tType = tTypes[0] #has to be one thing...exon or intron
#This only works because UTR takes precedence over EXON in TYPE.
if tType == 'EXON':
if codingTranscript:
codingFlag = 'C'
else:
codingFlag = 'NC'
else:
codingFlag = 'NC'
fOut.write('%s\t%s\t%s\t%s\t%s\t%s\n' % (eSite.ID, transcript.parent, transcript.id, tType, codingFlag, transcript.tType))
#fOut.write('%s:%s:%s\t%s\t%s\t%s\t%s\n' % (eSite.chromosome, eSite.strand, eSite.coordinate, transcript.parent, transcript.id, tType, codingFlag))
fOut.close()
def updateContext(oDir, geneSetFN):
print 'loading oRNA'
oDC = cgDB.dataController(oDir, cgOriginRNA.OriginRNA)
id_oRNA = oDC.load()
print 'loading gene set'
geneSet = cgGenes3.createGeneSetEditing(geneSetFN)
#Get in terms of tccs
print 'Joining'
oTcc_oRNA = oneToOne(id_oRNA.values(), 'tcc')
tTcc_transcripts = oneToMany(geneSet.transcripts, 'tcc')
#Overlap tccs
print 'overlapping'
oTcc_tTccs = compareData.getIndividualOverlaps(oTcc_oRNA.keys(), tTcc_transcripts.keys(), 1)
#create final dictionary containing {oRNA : [transcript, ..]}
oRNA_transcripts = {}
for oTcc in oTcc_tTccs:
oRNA = oTcc_oRNA[oTcc]
oRNA_transcripts[oRNA] = []
for tTcc in oTcc_tTccs[oTcc]:
oRNA_transcripts[oRNA].extend(tTcc_transcripts[tTcc])
print 'get context info'
#Go through each site and find out what it overlaps, and if it is in a coding region...
ds = bioLibCG.dominantSpotter(['EXON_INTRON', '3UTR', '5UTR', 'EXON', 'INTRON'])
for oRNA in oRNA_transcripts:
oRNA.transcriptIDs = []
oRNA.transcriptContexts = []
oRNA.transcriptTypes = []
oRNA.transcriptCodingTypes = []
if len(oRNA_transcripts[oRNA]) == 0:
continue
for transcript in oRNA_transcripts[oRNA]:
codingTranscript = '_coding' in transcript.tType
tType = None
codingFlag = None
tTypes = [x[1] for x in transcript.getOverlappingElements(oRNA.tcc)]
#categorize border types
tType = ds.spotItem(tTypes)
if tType == 'EXON' or 'EXON_INTRON':
if codingTranscript:
codingFlag = 'C'
else:
codingFlag = 'NC'
else:
codingFlag = 'NC'
oRNA.transcriptIDs.append(transcript.id)
oRNA.transcriptContexts.append(tType)
oRNA.transcriptTypes.append(transcript.tType)
oRNA.transcriptCodingTypes.append(codingFlag)
oDC.commit(id_oRNA)
def checkSeeds(editFN, contextFN, miLocationFN, miSequenceFN, gFN):
eSites = cgEdit.loadEditingSites(editFN)
cgEdit.updateContextEditingSites(eSites, contextFN) #puts the UTR, EXON in eSite.context
geneSet = cgGenes3.createGeneSetEditing(gFN)
tName_t = {}
for t in geneSet.transcripts:
tName_t[t.id] = t
miName_miSequence = {}
f = open(miSequenceFN, 'r')
for line in f:
ls = line.strip().split('\t')
name = ls[0]
seq = ls[1]
name = 'hsa-' + name
miName_miSequence[name] = seq
tName_miInfo = {}
f = open(miLocationFN, 'r')
for line in f:
ls = line.strip().split('\t')
tName = ls[0]
miName = ls[1]
loc = int(ls[2])
tName_miInfo.setdefault(tName, []).append([miName, loc])
foundIt = []
notFoundIt = []
for tName in tName_miInfo:
try:
t = tName_t[tName]
except:
continue
checkSeq = get3UTRSeq(t)
try:
mRNA = t.getMRNA()
except:
continue
for miInfo in tName_miInfo[tName]:
miName = miInfo[0]
loc = miInfo[1]
try:
miSequence = miName_miSequence[miName]
miSeed = miSequence[1:8]
except:
continue
rcMiSeed = cgSeqMod.reverseComplementSequence(miSeed, True)
newLoc = loc - (len(mRNA) - len(checkSeq))
finding = checkSeq.find(rcMiSeed, newLoc - 25)
if finding != -1:
if (0 < newLoc - finding < 30):
newResult = '%s\t%s\t%s\t%s\t%s' % (miName, tName, finding, newLoc, loc)
if newResult not in foundIt: foundIt.append(newResult)
else:
if miName == 'hsa-miR-21':
print loc, len(checkSeq), len(mRNA)
print mRNA
print checkSeq
newResult = '%s\t%s\t%s\t%s\t%s' % (miName, tName, finding, newLoc, loc)
if newResult not in notFoundIt: notFoundIt.append(newResult)
print len(foundIt)
print len(notFoundIt)
print ''
for i in foundIt:
print i
print ''
for i in notFoundIt:
print i
def makeContextPieBetter(contextFN, gFN, eFN, passedInfo, outFN):
if passedInfo[0] == passedInfo[1]:
print passedInfo, 'no such thing'
return 0
print 'loading geneSet'
geneSet = cgGenes3.createGeneSetEditing(gFN)
typeCount = {
'EXON': 0,
'3UTR': 0,
'5UTR': 0,
'INTRON': 0,
'NONG': 0,
'NONT': 0
}
#joint
tID_transcript = {}
for transcript in geneSet.transcripts:
tID_transcript[transcript.id] = transcript
eID_Info = {}
f = open(eFN, 'r')
for line in f:
ls = line.strip().split('\t')
eID = int(ls[13])
chrom = ls[0]
coord = ls[1]
newCoord = '%s:%s' % (chrom, coord)
nEdited = ls[4]
nTotal = ls[5]
eID_Info[eID] = [newCoord, nEdited, nTotal]
eID_gName = {}
eID_tTypes = {}
f = open(contextFN, 'r')
for line in f:
ls = line.strip().split('\t')
eID = int(ls[0])
gName = ls[1]
eID_gName[eID] = gName
tType = ls[3]
tName = ls[2]
if tType == 'INTER':
continue
transcript = tID_transcript[tName]
tInfo = [tType, transcript]
if eID in eID_tTypes:
eID_tTypes[eID].append(tInfo)
else:
eID_tTypes[eID] = [tInfo]
eID_finalType = {}
for eID in eID_tTypes:
highestType = None
for tInfo in eID_tTypes[eID]:
tType = tInfo[0]
transcript = tInfo[1]
tCoding = True
if '_coding' not in transcript.tType:
tCoding = False
gCoding = True
if '_coding' not in transcript.gType:
gCoding = False
#print transcript.id, tType, tCoding, gCoding
if tType == 'EXON':
if tCoding:
highestType = 'EXON'
break
else:
if gCoding:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR'
]:
highestType = 'NONT'
else:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR', 'NONT'
]:
highestType = 'NONG'
elif tType == '3UTR':
if tCoding:
if gCoding:
highestType = '3UTR'
else:
highestType = '3UTR'
else:
if gCoding:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR'
]:
highestType = 'NONT'
else:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR', 'NONT'
]:
highestType = 'NONG'
elif tType == '5UTR':
if tCoding:
if gCoding:
if highestType not in ['3UTR']:
highestType = '5UTR'
else:
if highestType not in ['3UTR']:
highestType = '5UTR'
else:
if gCoding:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR'
]:
highestType = 'NONT'
else:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR', 'NONT'
]:
highestType = 'NONG'
elif tType == 'INTRON':
if tCoding:
if gCoding:
if highestType not in ['3UTR', '5UTR']:
highestType = 'INTRON'
else:
if highestType not in ['3UTR', '5UTR']:
highestType = 'INTRON'
else:
if gCoding:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR'
]:
highestType = 'NONT'
else:
if highestType not in [
'EXON', 'INTRON', '3UTR', '5UTR', 'NONT'
]:
highestType = 'NONG'
eID_finalType[eID] = highestType
typeCount[highestType] += 1
outF = open(outFN, 'w')
for eID in eID_finalType:
outF.write('%s\t%s\t%s\t%s\t%s\t%s\t%s\n' %
(eID_Info[eID][0].split(':')[0],
eID_Info[eID][0].split(':')[1], passedInfo, eID_gName[eID],
eID_finalType[eID], eID_Info[eID][1], eID_Info[eID][2]))
return 0
outF = open(outFN, 'w')
for type in typeCount:
outF.write('%s\t%s\n' % (type, typeCount[type]))
return 0
#get fractions of each type
types = ['EXON', '3UTR', '5UTR', 'INTRON', 'NONG', 'NONT']
fracs = [
typeCount['EXON'], typeCount['3UTR'], typeCount['5UTR'],
typeCount['INTRON'], typeCount['NONG'], typeCount['NONT']
]
#print fracs
labels = [
'Exons (%s)' % fracs[0],
'3\'UTR (%s)' % fracs[1],
'5\'UTR (%s)' % fracs[2],
'Introns (%s)' % fracs[3],
'Noncoding Gene (%s)' % fracs[4],
'Noncoding Transcript (%s)' % fracs[5]
]
theSum = fracs[0] + fracs[1] + fracs[2] + fracs[3] + fracs[4] + fracs[5]
fracs = [float(x) / theSum for x in fracs]
#print fracs
explode = (0.1, 0.1, 0.1, 0.1, .1, .1)
pie(fracs, explode=explode, labels=labels, autopct='%1.1f%%', shadow=True)
title('Editing Site Genomic Location',
bbox={
'facecolor': '1.0',
'pad': 10
})
show()
def updateSynonomous(eFN, gFN, resultsFN, outFN):
#Load Transcripts and Editing Sites
print 'Loading editing sites'
eSites = cgEdit.loadEditingSites(eFN)
print 'Loading gene set'
geneSet = cgGenes3.createGeneSetEditing(gFN)
codingTID_eID = {}
f = open(resultsFN, 'r')
for line in f:
ls = line.strip().split('\t')
if ls[4] == 'C':
codingTID_eID[ls[2]] = int(ls[0])
#Get coding Transcripts
codingTranscripts = {} #tID : eID ! many:one always!
f = open(resultsFN, 'r')
for line in f:
ls = line.strip().split('\t')
if ls[4] == 'C':
codingTranscripts[ls[2]] = int(ls[0])
eID_eSite = {}
for eSite in eSites:
eID_eSite[eSite.ID] = eSite
tID_transcript = {}
for transcript in geneSet.transcripts:
tID[transcript.id] = transcript
codingT_eSite
for tID in codingTID_eID:
eID = codingTID_eID[tID]
t = tID_transcript[tID]
e = eID_eSite[eID]
print 'Creating scroll dict'
scrollDict = {} # transcript: eSite
for tID in codingTranscripts:
e = eJoint[codingTranscripts[tID]]
try:
t = tJoint[tID]
scrollDict[t] = e
except KeyError:
pass
print 'Deducing synonomous'
map = cgSeqMod.loadCodonMap('hg19')
finalDict = {} # tID: [SYN, AAA, AAB, G, A]
#Figure out if they are synonomous
for t in scrollDict:
eSite = scrollDict[t]
#dumpObj.dumpObj(t)
#dumpObj.dumpObj(eSite)
ePositionInMRNA = t.getRelativePositionMRNA(eSite.coordinate - 1)
if ePositionInMRNA == -1:
print t.id, 'should not be designated coding...'
continue
#grab mRNA and emRNA
mRNA = t.getMRNA(coding=True)
emRNA = t.getMRNA(coding=True)
if mRNA[ePositionInMRNA] != 'A':
print 'wrong position', t.id, '%s:%s' % (
eSite.chromosome, eSite.coordinate), eSite.strand, mRNA[
ePositionInMRNA - 5:ePositionInMRNA -
1], mRNA[ePositionInMRNA], mRNA[ePositionInMRNA +
1:ePositionInMRNA + 5]
#edit the site
emRNA = list(emRNA)
emRNA[ePositionInMRNA] = 'G'
emRNA = ''.join(emRNA)
#Test the protein sequences
pRNA = cgSeqMod.translateRNA(mRNA, map)
epRNA = cgSeqMod.translateRNA(emRNA, map)
#print t.parent, t.id
newString = ['%s ' % x for x in list(pRNA)]
newString = ''.join(newString)
if pRNA[0] != 'M':
print 'Non-canonical Start AA:', pRNA[0:5], mRNA[:10]
if pRNA[-1] != '*':
print 'Non-canonical End AA:', pRNA[-5:], mRNA[-10:]
#compare the codons.
mCodonList = cgSeqMod.getCodonListFromRNA(mRNA)
emCodonList = cgSeqMod.getCodonListFromRNA(emRNA)
compareList = zip(mCodonList, emCodonList)
synFlag = 'SYN'
codonNumber = ePositionInMRNA // 3
codonPair = compareList[codonNumber]
print t.id
print eSite.ID
print mCodonList[:codonNumber]
print mRNA[:ePositionInMRNA]
bCodon = codonPair[0]
aCodon = codonPair[1]
baa = cgSeqMod.translateRNA(bCodon, map)
aaa = cgSeqMod.translateRNA(aCodon, map)
if baa != aaa:
synFlag = 'NON'
bCodonList = list(bCodon)
aCodonList = list(aCodon)
matchedLetters = zip(bCodonList, aCodonList)
for pair in matchedLetters:
if pair[0] != 'A':
if pair[1] == 'G' and pair[0] != 'G':
print 'messed up codon switch', bCodonList, aCodonList
print t.parent, '%s:%s' % (
eSite.chromosome, eSite.coordinate
), eSite.strand, bCodon, aCodon, baa, aaa
else:
synFlag = 'SYN'
finalDict[t.id] = [synFlag, bCodon, aCodon, baa, aaa]
print 'writing to file'
#update line by line
newLines = []
f = open(resultsFN, 'r')
for line in f:
newLine = line.strip()
tID = line.strip().split('\t')[2]
if tID in finalDict:
newLine = newLine + '\t%s\t%s\t%s\t%s\t%s\n' % (
finalDict[tID][0], finalDict[tID][1], finalDict[tID][2],
finalDict[tID][3], finalDict[tID][4])
else:
newLine = newLine + '\tNA\tNA\tNA\tNA\tNA\n'
newLines.append(newLine)
f.close()
#update file
f = open(outFN, 'w')
f.writelines(newLines)
f.close()
def betterSynonymous(eFN, gFN, contextFN, outFN, refBase='A', eBase='G'):
print 'loading e sites'
eSites = cgEdit.loadEditingSites(eFN)
print 'loading geneSet'
geneSet = cgGenes3.createGeneSetEditing(gFN)
contextInfo = {} # eID: tID : [UTR, C]
f = open(contextFN, 'r')
for line in f:
ls = line.strip().split('\t')
eID = int(ls[0])
tID = ls[2]
cInfo = [ls[3], ls[4]]
if eID not in contextInfo:
contextInfo[eID] = {}
contextInfo[eID][tID] = cInfo
else:
contextInfo[eID][tID] = cInfo
eID_tIDs = {}
f = open(contextFN, 'r')
for line in f:
ls = line.strip().split('\t')
eID = int(ls[0])
tID = ls[2]
if tID not in eID_tIDs.setdefault(eID, []): eID_tIDs[eID].append(tID)
eID_eSite = {}
for eSite in eSites:
eID_eSite[eSite.ID] = eSite
tID_transcript = {}
for transcript in geneSet.transcripts:
tID_transcript[transcript.id] = transcript
eSite_transcripts = {}
for eID in eID_tIDs:
eSite = eID_eSite[eID]
tList = []
for tID in eID_tIDs[eID]:
if tID == 'NONE': continue
if tID_transcript.get(tID, None) == None: continue
tList.append(tID_transcript[tID])
eSite_transcripts[eSite] = tList
outF = open(outFN, 'w')
map = cgSeqMod.loadCodonMap('hg19')
for eSite in eSite_transcripts:
for transcript in eSite_transcripts[eSite]:
siteType, codingType = contextInfo[eSite.ID][transcript.id]
if '_noncoding' in transcript.tType:
continue
if codingType != 'C':
continue
ePositionInMRNA = transcript.getRelativePositionMRNA(
eSite.coordinate - 1)
mRNA = transcript.getMRNA(coding=True)
emRNA = transcript.getMRNA(coding=True)
if mRNA[ePositionInMRNA] != refBase:
print 'Editing site was not an A...'
#edit the site
emRNA = list(emRNA)
emRNA[ePositionInMRNA] = eBase
emRNA = ''.join(emRNA)
#Test the protein sequences
pRNA = cgSeqMod.translateRNA(mRNA, map)
epRNA = cgSeqMod.translateRNA(emRNA, map)
if pRNA[0] != 'M':
print 'Non-canonical Start AA:', pRNA[0:5], mRNA[:10]
if pRNA[-1] != '*':
print 'Non-canonical End AA:', pRNA[-5:], mRNA[-10:]
#compare the codons.
mCodonList = cgSeqMod.getCodonListFromRNA(mRNA)
emCodonList = cgSeqMod.getCodonListFromRNA(emRNA)
compareList = zip(mCodonList, emCodonList)
codonNumber = ePositionInMRNA // 3
codonPair = compareList[codonNumber]
bCodon = codonPair[0]
aCodon = codonPair[1]
baa = cgSeqMod.translateRNA(bCodon, map)
aaa = cgSeqMod.translateRNA(aCodon, map)
synFlag = 'SYN'
if baa != aaa:
synFlag = 'NON'
bCodonList = list(bCodon)
aCodonList = list(aCodon)
matchedLetters = zip(bCodonList, aCodonList)
for pair in matchedLetters:
if pair[0] != 'A':
if pair[1] == 'G' and pair[0] != 'G':
print 'messed up codon switch', bCodonList, aCodonList
print t.parent, '%s:%s' % (
eSite.chromosome, eSite.coordinate
), eSite.strand, bCodon, aCodon, baa, aaa
outF.write('\t'.join([
str(eSite.ID), transcript.parent, transcript.id, synFlag,
bCodon, aCodon, baa, aaa
]) + '\n')
def doit(fN):
geneSet = cgGenes3.createGeneSetEditing(fN)
for transcript in geneSet.transcripts:
if transcript.id == 'NM_031422':
dumpObj.dumpObj(transcript)
def makeTargetExpressionHistogram(eFN, targetFN, contextFN, geneFN, eChangeFN):
print 'loading expression ratios'
gName_eChange = getERatioDict(eChangeFN)
print 'loading eSites and Transcripts'
eSites = cgEdit.loadEditingSites(eFN)
geneSet = cgGenes3.createGeneSetEditing(geneFN)
print 'making joint dicts and loading extra data'
#joint
eID_eSite = {}
for eSite in eSites:
eID_eSite[eSite.ID] = eSite
#joint
tID_gName = {}
for transcript in geneSet.transcripts:
tID_gName[transcript.id] = transcript.parent
#load context data
f = open(contextFN, 'r')
eID_tID = {} # eID: tID
tID_tType = {}
for line in f:
ls = line.strip().split('\t')
eID = int(ls[0])
tID = ls[1]
tType = ls[2]
tID_tType[tID] = tType
if eID in eID_tID:
eID_tID[eID].append(tID)
else:
eID_tID[eID] = [tID]
f.close()
print 'analyzing'
#Get created or destroyed
f = open(targetFN, 'r')
altered = []
for line in f:
ls = line.strip().split('\t')
#created/destroyed
if (ls[3] != 'None') and (ls[4] == 'None'):
altered.append(int(ls[0]))
f.close()
print 'number of created/destroyed sites:', len(altered)
alteredSites = []
for id in altered:
alteredSites.append(eID_eSite[id])
eChanges = []
gDone = []
#Get gene names for each eSite
for eSite in alteredSites:
genes = []
for tID in eID_tID[eSite.ID]:
if tID == 'NONE':
continue
gName = tID_gName[tID]
if tID_tType[tID] != '3UTR':
print 'Not 3UTR', tID
continue
if gName not in genes:
genes.append(gName)
if len(genes) > 1:
print 'more than one gene for eSite...', genes
continue
if gName in gDone:
continue
else:
gDone.append(gName)
#Now add expression to HistoGram List...
if gName in gName_eChange:
eChange = gName_eChange[gName]
else:
print 'gene not in expression list', gName
continue
eChange = gName_eChange[gName]
eChange = math.log(eChange, 2)
eChanges.append(eChange)
#Now plot the histogram
plt.hist(eChanges, 40)
plt.xlabel('log2(RPKM KD/ RPKM CONTROL)')
plt.ylabel('# Genes')
plt.show()
