def cutSequence(sequence, motif) :
"""Takes a Biopython sequence and a motif and return the sequences cut at this motif.
Input:
sequence = the sequence
motif = a string containing a '-' at the cut site ; e.g. 'g-aattc' for EcoRI
Output:
a list of Biopython sequences SeqRecord, with names corresponding to the input sequence, the cut and the location in the input sequence
"""
# clean the cutting motif
pattern=motif.replace('-', '').upper()
cutSite=motif.find('-')
if (cutSite==-1) :
cutSite=len(motif) # if no '-' is found the cut is at the end of the motif
# extract the sequence string
try :
sequenceString=sequence.seq.data.upper() # if sequence is a SeqRecord
name=sequence.name+'_'+motif+'_'
except :
try :
sequenceString=sequence.data.upper() # if sequence is a Seq
name=motif+'_'
except :
sequenceString=sequence.upper() # sequence should be a string
name=motif+'_'
# find the first occurrence
fragments=[]
nextSite=sequenceString.find(pattern)
lastCut=1
lastPosition=len(sequenceString)
# loop
while (nextSite>0) :
# while a site is found
newFragment=SeqRecord(sequenceString[:nextSite+cutSite])
newFragment.name=name+str(lastCut)+'_'+str(lastCut+nextSite+cutSite-1)
lastCut=lastCut+nextSite+cutSite
sequenceString=sequenceString[nextSite+cutSite:]
fragments.append(newFragment)
nextSite=sequenceString.find(pattern)
# add the remaining sequence
if (sequenceString!='') :
name=name+str(lastCut)+'_'+str(lastPosition)
lastFragment=SeqRecord(sequenceString)
lastFragment.name=name
fragments.append(lastFragment)
# return
return fragments
def constructRandomChunks(sequences, numChunks=10, chunksize=1000):
"""Construct random sequ"""
rndChunks = list()
mapping = dict()
for s in sequences:
for c in range(numChunks):
rndPos = random.randint(0, len(s) - chunksize)
rndSeq = s.seq[rndPos:rndPos + chunksize]
rndID = str(uuid.uuid1())
rndChunks.append(SeqRecord(rndSeq, rndID))
mapping[rndID] = s.description
random.shuffle(rndChunks)
return rndChunks, mapping
def cutSequence(sequence, motif):
"""Takes a Biopython sequence and a motif and return the sequences cut at this motif.
Input:
sequence = the sequence
motif = a string containing a '-' at the cut site ; e.g. 'g-aattc' for EcoRI
Output:
a list of Biopython sequences SeqRecord, with names corresponding to the input sequence, the cut and the location in the input sequence
"""
# clean the cutting motif
pattern = motif.replace('-', '').upper()
cutSite = motif.find('-')
if (cutSite == -1):
cutSite = len(
motif) # if no '-' is found the cut is at the end of the motif
# extract the sequence string
try:
sequenceString = sequence.seq.data.upper(
) # if sequence is a SeqRecord
name = sequence.name + '_' + motif + '_'
except:
try:
sequenceString = sequence.data.upper() # if sequence is a Seq
name = motif + '_'
except:
sequenceString = sequence.upper() # sequence should be a string
name = motif + '_'
# find the first occurrence
fragments = []
nextSite = sequenceString.find(pattern)
lastCut = 1
lastPosition = len(sequenceString)
# loop
while (nextSite > 0):
# while a site is found
newFragment = SeqRecord(sequenceString[:nextSite + cutSite])
newFragment.name = name + str(lastCut) + '_' + str(lastCut + nextSite +
cutSite - 1)
lastCut = lastCut + nextSite + cutSite
sequenceString = sequenceString[nextSite + cutSite:]
fragments.append(newFragment)
nextSite = sequenceString.find(pattern)
# add the remaining sequence
if (sequenceString != ''):
name = name + str(lastCut) + '_' + str(lastPosition)
lastFragment = SeqRecord(sequenceString)
lastFragment.name = name
fragments.append(lastFragment)
# return
return fragments
def main():
# Configuration
#Select the desired NCBI translation table
translationTable = 11
# Open the DNA sequence file and read the fasta sequences into a dictionary
if (len(argv) > 1):
dnaFileName = argv[1]
else:
dnaFileName = None
dnaSeqFile = fileinput.input(dnaFileName)
dnaSeqDict = SeqIO.to_dict(SeqIO.parse(dnaSeqFile, "fasta"))
# Translate the sequences
aaSeqRecords = []
for key in dnaSeqDict:
aaSeq = SeqRecord(dnaSeqDict[key].seq.translate(table=translationTable), id=key)
aaSeqRecords.append(aaSeq)
dnaSeqFile.close()
# Replace stop codons with X (unknown aa) so muscle doesn't drop them
for aaSeq in aaSeqRecords:
noStopCodonSeq = str(aaSeq.seq).replace('*', 'X')
aaSeq.seq = Seq(noStopCodonSeq)
# Align the aa sequences
commandLine = str(MuscleCommandline(seqtype='protein'))
childProcess = subprocess.Popen(commandLine, stdin=subprocess.PIPE, stdout=subprocess.PIPE, shell=(sys.platform!="win32")) #don't pipe stderr or muscle hangs
SeqIO.write(aaSeqRecords, childProcess.stdin, "fasta")
childProcess.stdin.close()
aaAlignment = AlignIO.read(childProcess.stdout, "fasta")
# Convert the aa alignment into a dna alignment
dnaAlignment = Alignment(Gapped(IUPAC.unambiguous_dna, "-"))
for taxon in aaAlignment:
aaCount = 0
dnaSeq = ''
for aaResidue in taxon.seq:
if (aaResidue == '-'):
dnaSeq = dnaSeq + '---'
else:
dnaSeq = dnaSeq + dnaSeqDict[taxon.id].seq[aaCount*3:aaCount*3+3]
aaCount+=1
# As we add the sequences to the alignment remove gene name from the sequence id so they taxon match the PAML constraint tree
dnaAlignment.add_sequence(taxon.id.split('_')[0], str(dnaSeq))
if (dnaFileName):
outFileName = dnaFileName.split('.')[0] + '_aln.phy'
else:
outFileName = 'out_aln.phy'
outFile = open(outFileName, 'w+')
AlignIO.write([dnaAlignment], outFile, "phylip")
#I think this section should be removed. If I put the 'I' into the alignment file now, I can't open the alignment with BioPython-based scripts (for manual editing etc). I can use pamlize.py to add the I right before using paml.
# Biopython doesn't tag Interleaved phylip files and PAML requires it so...
# outFile.seek(0,0)
# modifiedAlignmentText = outFile.readlines()
# modifiedAlignmentText[0] = modifiedAlignmentText[0].rstrip() + ' I\n'
# outFile.seek(0,0)
# outFile.writelines(modifiedAlignmentText)
outFile.close()
alignmentNoGenSignalIterator = AlignIO.parse(fastaFileCladeNoGeneralSignal,"fasta",alphabet=Gapped(IUPAC.ExtendedIUPACProtein(),"-"));
alignmentIterator = AlignIO.parse(fastaFileClade,"fasta",alphabet=Gapped(IUPAC.ExtendedIUPACProtein(),"-"));
noGenSignalAlignment = alignmentNoGenSignalIterator.next()
queryFasta = resultFolder+"/"+"Query_%d.faa" % (entryToTest,)
ownCladeProfile = resultFolder+"/"+"ForOwnCladeProfile_%d.faa" % (entryToTest,)
#print testAlignment[entryToTest].id
#print testAlignment[entryToTest].seq
alignmentWithSignal = alignmentIterator.next()
desiredSeqString = str(alignmentWithSignal[entryToTest-1].seq)
desiredSeqString = desiredSeqString.replace("-", "")
#print desiredSeqString
seqNoGaps = Seq(desiredSeqString, alphabet=IUPAC.ExtendedIUPACProtein())
#print seqNoGaps
seqRecNoGaps = SeqRecord(seq=seqNoGaps, id=alignmentWithSignal[entryToTest-1].id)
#print seqRecNoGaps.seq
#print seqRecNoGaps.id
SeqIO.write(seqRecNoGaps, queryFasta, "fasta")
#print "Number of entries: ",len(testAlignment)
# Here we remove the desired element
newTestAlignment = []
for i in range(len(noGenSignalAlignment)):
if i != entryToTest-1 :
newTestAlignment.append(noGenSignalAlignment[i])
newAlignment = MultipleSeqAlignment(newTestAlignment)
AlignIO.write(newAlignment, ownCladeProfile,"fasta")
#print "Number of entries after: ",len(newTestAlignment)
