def primerStrip(file, GoodOut, BadOut, fwdprimer, revprimer):
PL = len(fwdprimer)
with open(GoodOut, 'w') as good:
with open(BadOut, 'w') as bad:
for title, seq, qual in FastqGeneralIterator(open(file)):
Diffs = primer.MatchPrefix(seq, fwdprimer)
if Diffs <= args.primer_mismatch:
Seq = seq[PL:]
Qual = qual[PL:]
if revprimer:#now need to look for reverse primer
BestPosRev, BestDiffsRev = primer.BestMatch2(Seq, revcomp_lib.RevComp(revprimer), args.primer_mismatch)
if BestPosRev > 0: #reverse primer was found
Seq = Seq[:BestPosRev]
Qual = Qual[:BestPosRev]
good.write("@%s\n%s\n+\n%s\n" % (title, Seq, Qual))
else:
bad.write("@%s\n%s\n+\n%s\n" % (title, seq, qual))
#this will loop through FASTQ file once, splitting those where barcodes are found, and primers trimmed
runningTotal = 0
trim = len(FwdPrimer)
#print Barcodes
with open(args.FASTQ, 'rU') as input:
for title, seq, qual in FastqGeneralIterator(input):
Barcode, BarcodeLabel = FindBarcode(seq, Barcodes)
if Barcode == "": #if not found, move onto next record
continue
BarcodeLength = len(Barcode)
seq = seq[BarcodeLength:]
qual = qual[BarcodeLength:]
#look for forward primer
if args.require_primer != 'off': #means we only want ones with forward primer and or reverse
Diffs = primer.MatchPrefix(seq, FwdPrimer)
if Diffs > args.primer_mismatch:
continue
#if found, trim away primer
seq = seq[trim:]
qual = qual[trim:]
if args.require_primer == 'both':
#look for reverse primer, strip if found
BestPosRev, BestDiffsRev = primer.BestMatch2(
seq, ReverseCompRev, args.primer_mismatch)
if BestPosRev > 0:
seq = seq[:BestPosRev]
qual = qual[:BestPosRev]
else:
continue
#check size
def processRead(input):
base = os.path.basename(input).split('.')[0]
PL = len(FwdPrimer)
RL = len(RevPrimer)
DemuxOut = os.path.join(tmpdir, base+'.demux.fq')
StatsOut = os.path.join(tmpdir, base+'.stats')
Total = 0
NoBarcode = 0
NoRevBarcode = 0
NoPrimer = 0
TooShort = 0
RevPrimerFound = 0
ValidSeqs = 0
with open(StatsOut, 'w') as counts:
with open(DemuxOut, 'w') as out:
for title, seq, qual in FastqGeneralIterator(open(input)):
Total += 1
#look for barcode, trim it off
Barcode, BarcodeLabel = FindBarcode(seq, Barcodes)
if Barcode == "":
NoBarcode += 1
continue
BarcodeLength = len(Barcode)
Seq = seq[BarcodeLength:]
Qual = qual[BarcodeLength:]
#now search for forward primer
Diffs = primer.MatchPrefix(Seq, FwdPrimer)
if Diffs > args.primer_mismatch:
NoPrimer += 1
continue
ForTrim = PL
#now search for reverse primer
BestPosRev, BestDiffsRev = primer.BestMatch2(Seq, RevPrimer, args.primer_mismatch)
if BestPosRev > 0: #reverse primer was found
RevPrimerFound += 1
#location to trim sequences
RevTrim = BestPosRev
#determine reverse barcode
if args.reverse_barcode:
RevBCdiffs = 0
BCcut = BestPosRev + RL
CutSeq = Seq[BCcut:]
RevBarcode, RevBarcodeLabel = FindBarcode(CutSeq, RevBarcodes)
if RevBarcode == "":
NoRevBarcode += 1
continue
BarcodeLabel = BarcodeLabel+'_'+RevBarcodeLabel
#now trim record remove forward and reverse reads
Seq = Seq[ForTrim:RevTrim]
Qual = Qual[ForTrim:RevTrim]
#since found reverse primer, now also need to pad/trim
if not args.full_length:
#check minimum length here or primer dimer type sequences will get padded with Ns
if len(Seq) < int(args.min_len):
TooShort += 1
continue
if len(Seq) < args.trim_len and args.pad == 'on':
pad = args.trim_len - len(Seq)
Seq = Seq + pad*'N'
Qual = Qual +pad*'J'
else: #len(Seq) > args.trim_len:
Seq = Seq[:args.trim_len]
Qual = Qual[:args.trim_len]
else:
#trim record, did not find reverse primer
if args.full_length: #if full length then move to next record
continue
#trim away forward primer
Seq = Seq[ForTrim:]
Qual = Qual[ForTrim:]
#check length and trim, throw away if too short as it was bad read
if len(Seq) < args.trim_len:
TooShort += 1
continue
Seq = Seq[:args.trim_len]
Qual = Qual[:args.trim_len]
#check minimum length
if len(Seq) < int(args.min_len):
TooShort += 1
continue
ValidSeqs += 1
#rename header
Name = 'R_'+str(ValidSeqs)+';barcodelabel='+BarcodeLabel+';'
out.write("@%s\n%s\n+\n%s\n" % (Name, Seq, Qual))
counts.write('%i,%i,%i,%i,%i,%i,%i\n' % (Total, NoBarcode, NoPrimer, RevPrimerFound, NoRevBarcode, TooShort, ValidSeqs))
def OnRec(Label, Seq, Qual):
global PL, LabelPrefix, SeqCount, OutCount, TooShortCount, PadCount
global FwdPrimerMismatchCount, RevPrimerStrippedCount
global FwdPrimer, RevPrimer
if SeqCount == 0:
progress.InitFile(fastq.File)
progress.File("%u reads, %u outupt, %u bad fwd primer, %u rev primer stripped, %u too short. %u padded" % \
(SeqCount, OutCount, FwdPrimerMismatchCount, RevPrimerStrippedCount, TooShortCount, PadCount))
SeqCount += 1
Seq = Seq
Qual = Qual
Diffs = MatchesPrimer(Seq, FwdPrimer)
if Diffs > MAX_PRIMER_MISMATCHES:
FwdPrimerMismatchCount += 1
return
OutCount += 1
Label = LabelPrefix + str(OutCount) + ";barcodelabel=" + SampleLabel + ";"
# Strip fwd primer
Seq = Seq[PL:]
Qual = Qual[PL:]
BestPosRev, BestDiffsRev = primer.BestMatch2(Seq, RevPrimer,
MAX_PRIMER_MISMATCHES)
if BestPosRev > 0:
# Strip rev primer
RevPrimerStrippedCount += 1
StrippedSeq = Seq[:BestPosRev]
StrippedQual = Qual[:BestPosRev]
# correctness checks
if 1:
Tail = Seq[BestPosRev:]
Diffs2 = primer.MatchPrefix(Tail, RevPrimer)
if Diffs2 != BestDiffsRev:
print >> sys.stderr
print >> sys.stderr, " Seq=" + Seq
print >> sys.stderr, "Tail=" + Tail
print >> sys.stderr, "RevP=" + RevPrimer
die.Die("BestPosRev %u Diffs2 %u BestDiffsRev %u" %
(BestPosRev, Diffs2, BestDiffsRev))
assert StrippedSeq + Tail == Seq
Seq = StrippedSeq
Qual = StrippedQual
L = len(Seq)
assert len(Qual) == L
if L < MinLen:
return
if L < TrimLen:
PadCount += 1
Seq = Seq + (TrimLen - L) * 'N'
Qual = Qual + (TrimLen - L) * 'I'
L = len(Seq)
assert L == TrimLen
assert len(Qual) == TrimLen
L = len(Seq)
if L < TrimLen:
TooShortCount += 1
return
if L > TrimLen:
Seq = Seq[:TrimLen]
Qual = Qual[:TrimLen]
L = len(Seq)
assert L == TrimLen
assert len(Qual) == TrimLen
fastq.WriteRec(out_file, Label, Seq, Qual)
def MatchesPrimer(Seq, Primer):
return primer.MatchPrefix(Seq, Primer)
def processRead(input):
#input is expected to be a FASTQ file
#local variables that need to be previously declared: ForPrimer, RevPrimer
Name = os.path.basename(input).split(".fq",-1)[0]
DemuxOut = os.path.join(args.out, Name + '.demux.fq')
Sample = Name.split('_')[0]
StatsOut = os.path.join(args.out, Name+'.stats')
Total = 0
NoPrimer = 0
TooShort = 0
RevPrimerFound = 0
ValidSeqs = 0
PL = len(FwdPrimer)
with open(StatsOut, 'w') as counts:
with open(DemuxOut, 'w') as out:
for title, seq, qual in FastqGeneralIterator(open(input)):
Total += 1
#first thing is look for forward primer, if found trim it off
Diffs = primer.MatchPrefix(seq, FwdPrimer)
#if require primer is on make finding primer in amplicon required if amplicon is larger than read length
#if less than read length, can't enforce primer because could have been trimmed via staggered trim in fastq_mergepairs
if args.primer == 'on' and len(seq) > ReadLen:
if Diffs > args.primer_mismatch:
NoPrimer += 1
continue
Seq = seq[PL:]
Qual = qual[PL:]
else:
if Diffs <= args.primer_mismatch:
Seq = seq[PL:]
Qual = qual[PL:]
else:
NoPrimer += 1
Seq = seq
Qual = qual
#now look for reverse primer
BestPosRev, BestDiffsRev = primer.BestMatch2(Seq, RevPrimer, args.primer_mismatch)
if BestPosRev > 0: #reverse primer was found
RevPrimerFound += 1
#location to trim sequences, trim seqs
Seq = Seq[:BestPosRev]
Qual = Qual[:BestPosRev]
else:
if args.full_length and len(Seq) > ReadLen: #if full length and no primer found, exit, except when length is less than read length
continue
#if full_length is passed, then only trim primers
if not args.full_length:
#got here if primers were found they were trimmed
#now check seq length, pad if too short, trim if too long
if len(Seq) < args.min_len: #need this check here or primer dimers will get through
TooShort += 1
continue
if len(Seq) < args.trim_len and args.pad == 'on':
pad = args.trim_len - len(Seq)
Seq = Seq + pad*'N'
Qual = Qual +pad*'J'
else: #len(Seq) > args.trim_len:
Seq = Seq[:args.trim_len]
Qual = Qual[:args.trim_len]
#got here, reads are primers trimmed and trim/padded, check length
if len(Seq) < args.min_len:
TooShort += 1
continue
ValidSeqs += 1
#now fix header
Title = 'R_'+str(ValidSeqs)+';barcodelabel='+Sample+';'
#now write to file
out.write("@%s\n%s\n+\n%s\n" % (Title, Seq, Qual))
counts.write('%i,%i,%i,%i,%i\n' % (Total, NoPrimer, RevPrimerFound, TooShort, ValidSeqs))