def collectMiRNA_totals(pathToFile):
lines = open(pathToFile,'rU').readlines()
groupedLines = groupByField_silent(lines,0,sep=' : ')
groupedLines.sort(key=lambda x: x[0][0])
rDict = {}
for miR in groupedLines:
data = Bag({'name':miR[0][0],
'orthoTypes':[],
'AGAPs':initList(4,set())
})
for line in miR:
if line[1].startswith("allPassedSeedsFor_"):
orthoType = int(line[1][-1])
data.orthoTypes.append(orthoType)
agaps = []
for group in eval(line[-1]):
for gene in group:
if gene.startswith('AGAP'):
agaps.append(gene)
data.AGAPs[orthoType].update(agaps)
rDict[data.name]=data
return rDict
def collectGeneNames(goTerm,masterGenes,anoXdict):
"""\tFor a GO term, query anoXdict for which genes in masterGenes(real & ctrls)
are tagged with itself, and return a list([ set(real) , [sets(ctrl)] ]).
RETURNS: a list([ set(realGenes) , [sets(ctrlGenes)] ])
"""
rList = [set(),initList(ctrlNum,set())]
notInAnoXcel = set()
# -- Collect Reals --
"""Should Collect those gene Names not found in AnoXcel"""
for gene in masterGenes[0]:
try:
anoXdata[gene]
except KeyError:
notInAnoXcel.add(gene)
continue
geneData = [x[-2] for x in anoXdata[gene]]
if goTerm in geneData:
rList[0].add(gene)
# -- Collect Ctrls --
"""Should Collect those gene Names not found in AnoXcel"""
for i in range(ctrlNum):
for gene in masterGenes[1][i]:
try:
anoXdata[gene]
except KeyError:
notInAnoXcel.add(gene)
continue
geneData = [x[-2] for x in anoXdata[gene]]
if goTerm in geneData:
rList[1][i].add(gene)
# Return
return rList
else:
anoXdata[row[0][:-3]] = [[row[0]]+row[strt:stp]]
goInfo[row[strt:stp][-2]] = row[strt:stp] # Create a library of goTermInfo
# -- Good miRNA A1-m8 list --
print "processing Good miRNA A1-m8 list..."
A1_m8s = sorted(map(lambda l: l.strip('\n'), open(A1_m8_file,'rU')))
# -- Process Events Pickle --
print "processing Events Pickle..."
data_Ca = cPickle.load(open(pklPath_Ca,'rU'))
classConvert = {'II':2,'III':3}
ctrlNum = len(data_Ca[data_Ca.keys()[0]].ctrlEvents['A1_to_m8'])
masterGeneSets = [set(),initList(ctrlNum,set())]
getMasterTargetList(useClass,A1_m8s,masterGeneSets, data_Ca)
print 'deleting Events Pickle...'
del(data_Ca) # free-up some memory to work with!
# --------- Main Body ---------
# -- Collect GO term Data --
print 'Collecting GO-term Data...'
def writeTargetsFdrMedMeth(miRobj_Ca,miRobj_Cb,oFile):
#print 'Processing %s...' % (miRobj_Ca.name)
miRHits_Ca = miRobj_Ca.reportGeneTargetsFdrMedMeth(stdvLimit=stdvsAboveMed,
consFdrThresh=consFdrThreshold,
divide=0)
totReal = [None,set(),set(),set()]
totCtrl = initList(len(miRobj_Ca.ctrlEvents[miRobj_Ca.ctrlEvents.keys()[0]]),[None, set(), set(),set()])
# Calulate combined FDR for miRNA using Ctrl_b data from seedTypes that passed
# the reportGeneTargetsFdrMedMeth() Ctrl_a screen.
# >> Gather and combine data from passed seedTypes:
for oType in range(1,4):
for sType in miRT._seedModels:
if miRHits_Ca[sType][oType]:
trLen_0 = len(totReal[oType])
totReal[oType].update(miRHits_Ca[sType][oType][0])
rUpdtLen = len(miRHits_Ca[sType][oType][0])
trLen_1 = len(totReal[oType])
None
for i in range(len(totCtrl)):
tciLen_0 = len(totCtrl[i][oType])
totCtrl[i][oType].update(miRobj_Cb.ctrlEvents[sType][i][oType])
cUpdtLen = len(miRobj_Cb.ctrlEvents[sType][i][oType])
tciLen_1 = len(totCtrl[i][oType])
None
# >> Calculate separate FDRs for each Ctrl_b group:
totalsData = [None,None,None,None]
for oType in range(1,4):
if totReal[oType] == set():
continue
tempFDRs = []
for i in range(len(totCtrl)):
ctrlVal = len(totCtrl[i][oType])
realVal = len(totReal[oType])
if float(ctrlVal)/realVal >= 1:
tempFDRs.append(1.0)
else:
tempFDRs.append(float(ctrlVal)/realVal)
tLen = len(tempFDRs)
oFDRstdv,oFDRmed = mathDefs.stdDv(tempFDRs,'median')
cons_oFDR = oFDRmed + (stdvsAboveMed*oFDRstdv)
totalsData[oType] = [totReal[oType],oFDRmed,cons_oFDR]
# Write out Totals data:
print miRobj_Cb.name
outFile.write('-- %s --\n' % (miRobj_Cb.name))
for i in range(1,len(totalsData)):
if totalsData[i]:
outFile.write('%s : allPassedSeedsFor_%s : %s : %s : %s Seqs=%s\n' \
%(miRobj_Cb.name,
i,
len(totalsData[i][0]),
totalsData[i][2],
totalsData[i][1],
','.join(sorted([str(x) for x in totalsData[i][0]]))))
# write out passed Seed data
for seedType in sorted(miRHits_Ca):
for i in range(1,len(miRHits_Ca[seedType])):
if miRHits_Ca[seedType][i]:
outFile.write('%s : %s : orthoType_%s : %s : %.2f : %.4f Seqs=%s\n'\
%(miRobj_Ca.name,
seedType,
i,
len(miRHits_Ca[seedType][i][0]),
miRHits_Ca[seedType][i][2],
miRHits_Ca[seedType][i][1],
','.join(sorted([str(x) for x in miRHits_Ca[seedType][i][0]]))))
outFile.flush()
