def nucTree(prog, opts, f, refSeq, wd, reftree, organelle):
orgGenomes = len(refSeq)
files = []
id = strftime("%H%M%S_nuc", gmtime())
revcmp = string.maketrans("ACGTNSWRYKMBDHVacgtnswrykmbdhv",
"TGCANSWYRMKVHDBtgcanswyrmkvhdb")
# generate multi-fasta sequence files
if opts.verbose:
print prog + ": create fasta file(s): insertions and corresponding organelle subsequences"
for name, seq in readFasta(f):
coord = name.split(":")
beg = map(int, coord[4].split(","))
end = map(int, coord[5].split(","))
refBeg, refEnd = findRefIndex(beg, end, refSeq[0][1])
if coord[-1] == "-":
seq = seq[::-1].translate(revcmp)
file = wd + "/" + name + ".mfa"
fp = open(file, "w")
chunk = len(refBeg)
writeFasta(">" + name, seq, fp)
for i in range(orgGenomes):
subseq = ""
for j in range(chunk):
subseq += refSeq[i][1][refBeg[j]:refEnd[j]]
if subseq:
writeFasta(">" + refSeq[i][0], subseq.replace("-", ""), fp)
fp.close()
files.append(file)
# compute multiple sequence alignments
flen = len(files)
tmp = wd + "/" + id
if opts.verbose:
print prog + ": compute multiple sequence alignments of insertion sequences"
for i in range(flen):
msa = files[i].split(".")[0] + ".msa"
exeMafft(opts.mafft, "--localpair", files[i], tmp, seqLen(files[i]))
fastaToPhylip(tmp, msa)
files[i] = msa
commands.getoutput("rm " + tmp)
# calculate phylogenetic trees
if opts.verbose:
print prog + ": calculate phylogenetic trees of insertion sequences"
for i in range(flen):
ptr = files[i].split(".")[0] + ".ptr"
exeRaxml(opts.raxml, files[i], ptr, id, wd, reftree, organelle)
files[i] = ptr
return files
def nucTree(prog, opts, f, refSeq, wd, reftree, organelle):
orgGenomes = len(refSeq)
files = []
id = strftime("%H%M%S_nuc", gmtime())
revcmp = string.maketrans("ACGTNSWRYKMBDHVacgtnswrykmbdhv",
"TGCANSWYRMKVHDBtgcanswyrmkvhdb")
# generate multi-fasta sequence files
if opts.verbose:
print prog + ": create fasta file(s): insertions and corresponding organelle subsequences"
for name, seq in readFasta(f):
coord = name.split(":")
beg = map(int, coord[4].split(","))
end = map(int, coord[5].split(","))
refBeg, refEnd = findRefIndex(beg, end, refSeq[0][1])
if coord[-1] == "-":
seq = seq[::-1].translate(revcmp)
file = wd + "/" + name + ".mfa"
fp = open(file, "w")
chunk = len(refBeg)
writeFasta(">" + name, seq, fp)
for i in range(orgGenomes):
subseq = ""
for j in range(chunk):
subseq += refSeq[i][1][refBeg[j]:refEnd[j]]
if subseq:
writeFasta(">" + refSeq[i][0], subseq.replace("-", ""), fp)
fp.close()
files.append(file)
# compute multiple sequence alignments
flen = len(files)
tmp = wd + "/" + id
if opts.verbose:
print prog + ": compute multiple sequence alignments of insertion sequences"
for i in range(flen):
msa = files[i].split(".")[0] + ".msa"
exeMafft(opts.mafft, "--localpair", files[i], tmp, seqLen(files[i]))
fastaToPhylip(tmp, msa)
files[i] = msa
commands.getoutput("rm " + tmp)
# calculate phylogenetic trees
if opts.verbose:
print prog + ": calculate phylogenetic trees of insertion sequences"
for i in range(flen):
ptr = files[i].split(".")[0] + ".ptr"
exeRaxml(opts.raxml, files[i], ptr, id, wd, reftree, organelle)
files[i] = ptr
return files
def extractBaitSeq(baitSeq, baitRev, genome, coord):
fs = open(baitSeq, "w")
fr = open(baitRev, "w")
for name, seq in readFasta(genome):
c = editBaitCoord(coord.get(name, []), 200)
if not c:
continue
for n, s in extractSeq(name, c, seq, True):
beg, end = n.split(":")[1:]
n = ">" + name + ":" + ",".join(beg.split(",")[1:-1]) + ":" + ",".join(end.split(",")[1:-1])
writeFasta(n, s, fs)
writeFasta(n, s[::-1], fr)
fs.close()
fr.close()
def extractBaitSeq(baitSeq, baitRev, genome, coord):
fs = open(baitSeq, "w")
fr = open(baitRev, "w")
for name, seq in readFasta(genome):
c = editBaitCoord(coord.get(name, []), 200)
if not c: continue
for n, s in extractSeq(name, c, seq, True):
beg, end = n.split(":")[1:]
n = ">" + name + ":" + \
",".join(beg.split(",")[1:-1]) + ":" + \
",".join(end.split(",")[1:-1])
writeFasta(n, s, fs)
writeFasta(n, s[::-1], fr)
fs.close()
fr.close()
def extractFlankSeq(flankSeq, flankRev, flanklen, hitDict, nucleus):
fs = open(flankSeq, "w")
fr = open(flankRev, "w")
wrote = 0
mark = string.maketrans(":", "^")
for name, seq in readFasta(nucleus):
coord = makeFlankCoord(hitDict.get(name, []), flanklen)
if not coord: continue
for n, s in extractSeq(name, coord, seq, True):
beg, end = n.split(":")[1:]
n = ">" + name.translate(mark) + ":" + \
beg.split(",")[-1] + ":" + end.split(",")[0]
if len(s):
writeFasta(n, s, fs)
writeFasta(n, s[::-1], fr)
wrote += 1
fs.close()
fr.close()
return wrote
def extractFlankSeq(flankSeq, flankRev, flanklen, hitDict, nucleus):
fs = open(flankSeq, "w")
fr = open(flankRev, "w")
wrote = 0
mark = string.maketrans(":", "^")
for name, seq in readFasta(nucleus):
coord = makeFlankCoord(hitDict.get(name, []), flanklen)
if not coord: continue
for n, s in extractSeq(name, coord, seq, True):
beg, end = n.split(":")[1:]
n = ">" + name.translate(mark) + ":" + \
beg.split(",")[-1] + ":" + end.split(",")[0]
if len(s):
writeFasta(n, s, fs)
writeFasta(n, s[::-1], fr)
wrote += 1
fs.close()
fr.close()
return wrote
def align(prog, opts, args, wd):
hitDict = {}
hitCounts = 0
# command line format strings
tantan = opts.tantan + "tantan %s > %s"
lastdb = opts.last + "lastdb %s %s %s"
lastal = opts.last + "lastal -e%s -j4 -f0 %s %s | grep -v '#'"
lastex = opts.last + "lastex -E%s %s.prj %s.prj | sed -n 4p - | cut -f1"
# (soft-masked) sequence file names
organelle = wd + "/orgSeq"
nucleus = wd + "/nucSeq"
# index file names
nucBaseFreq = wd + "/nucBaseFreq"
orgIndex = wd + "/orgIndex"
# simulation file names
orgRev = wd + "/orgRev"
orgRevIndex = wd + "/orgRevIndex"
# double circular organelle genome
if int(commands.getoutput("grep -c '>' " + args[0])) != 1:
raise Exception("there must be exactly 1 sequence: " + args[0])
commands.getoutput("cp " + args[0] + " " + organelle)
if opts.dnaform == 'circular':
if opts.verbose:
print prog + ": double organelle sequence"
commands.getoutput("grep -v '>' " + args[0] + " >> " + organelle)
# soft-mask genomes with 'tantan'
if opts.rmsk_organelle:
if opts.verbose:
print prog + ": soft-mask organelle genome"
tmpfile = wd + "/tmp"
commands.getoutput(tantan % (organelle, tmpfile))
commands.getoutput("mv " + tmpfile + " " + organelle)
if opts.rmsk_nucleus:
if opts.verbose:
print prog + ": soft-mask nuclear genome"
commands.getoutput(tantan % (args[1], nucleus))
else: nucleus = args[1]
# index genomes with 'lastdb'
if opts.rmsk_organelle or opts.rmsk_nucleus:
lastdbOption = " "
else:
lastdbOption = "-c"
commands.getoutput(lastdb % ("-x", nucBaseFreq, nucleus))
commands.getoutput(lastdb % (lastdbOption, orgIndex, organelle))
# calculate (and simulate) E-value
if opts.evalue > 0: opts.greedy = False
else : opts.evalue = 0.01
evalue = "%.2e" % opts.evalue
score = commands.getoutput(lastex % (evalue, orgIndex, nucBaseFreq))
if opts.greedy:
if opts.verbose:
print prog + ": simulate E-value threshold"
f = open(orgRev, "w")
for name, seq in readFasta(organelle):
writeFasta(">" + name, seq[::-1], f)
f.close()
commands.getoutput(lastdb % (lastdbOption, orgRevIndex, orgRev))
score, evalue = evalueSimulation(lastex, lastal, orgRevIndex,
nucBaseFreq, nucleus, score, evalue)
commands.getoutput("rm " + orgRev + " " + orgRevIndex + "*")
# align organelle and nuclear genomes
if opts.verbose:
print prog + ": set: e-value=" + evalue + " / score=" + score
print prog + ": start alignment"
alignResult = commands.getoutput(lastal % (score, orgIndex, nucleus))
if not alignResult:
raise Exception("no hits are found")
# format result
hitDict = formatResult(alignResult)
# merge overlapped hits
if opts.verbose:
print prog + ": cull overlapped alignments"
for chrom in hitDict.keys():
hitDict[chrom].sort()
hitDict[chrom] = mergeOverlap(hitDict[chrom])
hitCounts = sum(map(len, hitDict.values()))
return hitDict, hitCounts, organelle, nucleus
