def findSNRCutoff(self):
self.varsP.updateInfoReport( ' Finding SNR cutoff %s\n' % self.expTag, printalso=True ) #put this in info report
usedScanCount = 0
#print "\nnscans = ", len(self.scans), "\n" #debug
deviceDset = molecule.moleculeDataset(self.curExp.basesPerPixel,numLabelChannels=self.numLabelChannels)
self.varsP.updateInfoReport(" ID "+deviceDset.makeExperimentHeader(), printalso=True)
for i, scan in enumerate(self.scans):
#if i == 0:
#self.scanReportHeader += deviceDset.makeExperimentHeader() +'\n'
#if self.varsP.lambdaRef:
# deviceDsetLambda = molecule.moleculeDataset(self.curExp.basesPerPixel,numLabelChannels=1)
molFile = scan.molFile
if not os.path.exists(molFile) :
self.varsP.updatePipeReport("Device.findSNRCutoff: MISSING FILE: " + molFile + "\n")
continue
usedScanCount += 1
scanDset = molecule.moleculeDataset(self.curExp.basesPerPixel, molTag=int(scan.molTag),numLabelChannels=self.numLabelChannels)
scanDset.readMolFile(molFile)
lab2File = molFile.replace('.mol', '.0.lab')
scanDset.annotateLabels(lab2File)
if self.numLabelChannels > 1:
lab2File = molFile.replace('.mol', '.1.lab')
scanDset.annotateLabels(lab2File, labelChannel=1)
self.varsP.updateInfoReport( " SNR cutoff: " + scan.molTag + " " + scanDset.makeExperimentReport(), printalso=True)
deviceDset.addDset(scanDset)
if not usedScanCount :
return
self.roughMassProfileHeader = deviceDset.getRoughMassProfile(headerOnly=True)
self.roughMassProfile = deviceDset.getRoughMassProfile()
deviceDset.getLogSnrCutoff()
self.SnrCutoff = deviceDset.LogSnrSimpleThresh
self.lambdaSnrCutoff = self.SnrCutoff
self.varsP.updateInfoReport( " SNR total : " + deviceDset.makeExperimentReport(), printalso=True )
snrstr = " ".join( map(lambda x: "%.3f"%x, self.SnrCutoff) )
self.varsP.updateInfoReport( " SNR Cutoff: " + snrstr + "\n", printalso=True )
def getCorrectedDataset(self, bpp, snrCutoff, swapchannels=False, verbose=False):
'''Load mol and lab files into molecule.moleculeDataset,
call scaleTo500 and filteredSubset in order to apply SNR filter as previously calculated.'''
scanDset = molecule.moleculeDataset(bpp, molTag=int(self.molTag),numLabelChannels=self.numLabelChannels)
ret = scanDset.readMolFile(self.molFile)
if ret != 0 :
return
suff = '.0.lab'
if swapchannels :
suff = '.1.lab'
lab2File = self.molFile.replace('.mol', suff)
scanDset.annotateLabels(lab2File)
self.Occupancy = scanDset.CalculateOccupancy(1)
self.Mb = scanDset.MegabasesDetected
if verbose :
self.varsP.updateInfoReport(" Prefilter : %s %s" % (self.molTag, scanDset.makeExperimentReport()), printalso=True)
if self.numLabelChannels > 1:
lab2File = self.molFile.replace('.mol', '.1.lab')
scanDset.annotateLabels(lab2File, labelChannel=1)
scanDset.scaleTo500()
if verbose :
self.varsP.updateInfoReport(" Postscale : %s %s" % (self.molTag, scanDset.makeExperimentReport()), printalso=True) #this one is ok
filt = molecule.filteredSubset(scanDset,minLabSnr=snrCutoff,minLabels=5,minLen=self.minMolLen)
if verbose :
self.varsP.updateInfoReport(" Postfilter: %s %s" % (self.molTag, filt.makeExperimentReport()), printalso=True)
return filt
def getLambdaMapJob(self, snrCutoff=0, verbose=False):
#Note, verbose will print once per job, so use for debugging only
# add lambda alignment band to this
lambdaFilter = self.varsP.argData['lambdaFilter']
lamMinLen = float(lambdaFilter[lambdaFilter.index('-minlen') +1]) if '-minlen' in lambdaFilter else 40.
lamMaxLen = float(lambdaFilter[lambdaFilter.index('-maxlen') +1]) if '-maxlen' in lambdaFilter else 60.
lamMinLab = float(lambdaFilter[lambdaFilter.index('-minsites')+1]) if '-minsites' in lambdaFilter else 6.
lamMaxLab = float(lambdaFilter[lambdaFilter.index('-maxsites')+1]) if '-maxsites' in lambdaFilter else 10.
#old format below (dict, not list)
#lamMinLen = int(lambdaFilter['-minlen' ]) # 40
#lamMaxLen = int(lambdaFilter['-maxlen' ]) # 60
#lamMinLab = int(lambdaFilter['-minsites']) # 6
#lamMaxLab = int(lambdaFilter['-maxsites']) # 10
if verbose :
self.varsP.updateInfoReport("lamMinLen = %.0f\n" % lamMinLen, printalso=True)
self.varsP.updateInfoReport("lamMaxLen = %.0f\n" % lamMaxLen, printalso=True)
self.varsP.updateInfoReport("lamMinLab = %.0f\n" % lamMinLab, printalso=True)
self.varsP.updateInfoReport("lamMaxLab = %.0f\n" % lamMaxLab, printalso=True)
#need mol file to do this; if doesn't exist, return with warning
if not(os.path.exists(self.molFile)):
print "Skipping map lambda job", self.molTag, "because mol file missing:", self.molFile
self.lambdaErrFile = None
return
bnxFileLambda = '%s_lambda.bnx' % self.molTag
bnxFileLambda = os.path.join(os.path.split(self.molFile)[0], bnxFileLambda)
#if lambda bnx exists, skip the isolation step
if os.path.exists(bnxFileLambda) :
print "Using lambda bnx", bnxFileLambda
else :
print ' Isolating Lambda %s' % self.molTag
lab2File = self.molFile.replace('.mol', '.0.lab')
scanDset = molecule.moleculeDataset(self.curExp.basesPerPixel, molTag=int(self.molTag))
scanDset.readMolFile(self.molFile)
scanDset.annotateLabels(lab2File)
# Introduce optArguments for Lambda Band
scanDsetLambda = molecule.filteredSubset(scanDset,snrCutoff,lamMinLab,lamMaxLab,lamMinLen,lamMaxLen,True)
scanDsetLambda.writeBnxFile(bnxFileLambda, quality=self.quality)
self.lambdaBnx = bnxFileLambda
baseArgs = self.varsP.argsListed('mapLambda')
outputTarget = bnxFileLambda.replace('.bnx', '')
curArgs = [self.varsP.RefAlignerBin, '-i', bnxFileLambda, '-o', outputTarget, '-ref', self.varsP.lambdaRef] + baseArgs
if self.varsP.stdoutlog :
curArgs.extend( ['-stdout', '-stderr'] )
jobTag = self.molTag + '_lambda'
self.lambdaErrFile = outputTarget + '.err'
#if the err file exists, no need to process
if os.path.exists(self.lambdaErrFile) :
print "Skipping map lambda job ", jobTag, "because err file exists", self.lambdaErrFile
return
return mthread.singleJob(curArgs, jobTag, self.lambdaErrFile, jobTag)
def writeCorrectedBnx(self):
self.varsP.updateInfoReport(" Correcting %s\n" % self.expTag, printalso=True)
deviceDset = molecule.moleculeDataset(self.curExp.basesPerPixel,numLabelChannels=self.numLabelChannels)
self.varsP.updateInfoReport(" ID "+deviceDset.makeExperimentHeader(), printalso=True)
goodscans = 0
for i,scan in enumerate(self.scans):
#print scan.localTiff #debug
#print self.bppCorrection[i] #debug
#if self.bppCorrection != 1.0 :
correctedBpp = self.bppCorrection[i]*self.curExp.basesPerPixel
curDataset = scan.getCorrectedDataset(correctedBpp, self.SnrCutoff)
if not curDataset :
continue
goodscans += 1
#this is redundant with the info report at the end of performImageAnalysis
# partially, since that comes from the scan object, this is from the moleculeDataset
self.varsP.updateInfoReport(' Corrected : %s %s' % (scan.molTag, curDataset.makeExperimentReport()), printalso=True)
#self.varsP.updateInfoReport('\n', printalso=True) #debug
deviceDset.addDset(curDataset)
if not goodscans :
self.bnxFile = "" #if no good scans, null the bnxFile
return
self.labPer100kb = deviceDset.calculateLabPer100Kb()
deviceDset.writeBnxFile(self.bnxFile, quality=self.quality)