def raxTree(seqBlock, indNames, model, raxml, outgroup = None, uniqueTag = "", test = False, log="/dev/null"):
#write file
tempAln = open("temp." + uniqueTag + ".phy", "w")
tempAln.write(genomics.makeAlnString(indNames,seqBlock))
tempAln.close()
if outgroup is not None:
og = " -o " + ",".join(outgroup)
else:
og = ""
#raxCommand = raxml + " -s temp." + uniqueTag + ".phy -n " + uniqueTag + " -m " + model + og + " -V -f d -p 12345 --silent"
raxCommand = raxml + " -s temp." + uniqueTag + ".phy -n " + uniqueTag + " -m " + model + og + " -V -f d -p 12345 --silent >>" + log
if test: print >> sys.stderr, "raxml command:\n", raxCommand
os.system(raxCommand)
tempAln.close()
#try retrieve the result
try:
treeFile = open("RAxML_bestTree." + uniqueTag, "r")
tree = treeFile.readline()
treeFile.close()
except:
tree = "NA\n"
#remove files
if not test:
os.system("rm temp." + uniqueTag + ".phy*")
os.system("rm RAxML*" + uniqueTag)
return tree
def phymlTree(seqArray, seqNames, model, opt, phyml, prefix = "", tmpDir = None, test = False, log="/dev/null"):
#write file
tempAln = tempfile.NamedTemporaryFile(mode="w",prefix=prefix,suffix=".phy",dir=tmpDir,delete=False)
with tempAln as tA: tA.write(genomics.makeAlnString(seqNames,seqArray))
phymlCommand = " ".join([phyml,"--input", tempAln.name,"--model", model, "-o", opt, "-b 0", ">>", log])
if test: print >> sys.stderr, "phyml command:\n", phymlCommand
os.system(phymlCommand)
#try retrieve the result
try:
with open(tempAln.name + "_phyml_tree.txt", "r") as treeFile: tree = treeFile.readline().strip()
except:
try:
with open(tempAln.name + "_phyml_tree", "r") as treeFile: tree = treeFile.readline().strip()
except:
if verbose:
sys.stderr.write("Tree not found at " + tempAln.name + "_phyml_tree.txt\n")
tree = "NA"
try:
with open(tempAln.name + "_phyml_stats.txt", "r") as statsFile:
stats = statsFile.read().split()
lnL = stats[stats.index("Log-likelihood:")+1]
except:
try:
with open(tempAln.name + "_phyml_stats", "r") as statsFile:
stats = statsFile.read().split()
lnL = stats[stats.index("Log-likelihood:")+1]
except:
lnL = "NA"
#remove files
if not test: os.system("rm " + tempAln.name + "*")
return (tree,lnL,)
def phymlCrossVal(seqArray0, seqArray1, indNames, model, opt, phyml, prefix = "",tmpDir=None, test = False, log="/dev/null"):
#write file
tempAln0 = tempfile.NamedTemporaryFile(mode="w",prefix=prefix,suffix=".0.phy",dir=tmpDir,delete=False)
tempAln1 = tempfile.NamedTemporaryFile(mode="w",prefix=prefix,suffix=".1.phy",dir=tmpDir,delete=False)
localName0 = tempAln0.name.rsplit("/",1)[1]
localName1 = tempAln1.name.rsplit("/",1)[1]
with tempAln0 as tempAln0: tempAln0.write(genomics.makeAlnString(seqNames,seqArray0))
with tempAln1 as tempAln1: tempAln1.write(genomics.makeAlnString(seqNames,seqArray1))
#first way validation
#tree
phymlCommand = " ".join([phyml,"--input", tempAln0.name,"--model", model, "-o", opt, ">>", log])
os.system(phymlCommand)
#validation
phymlCommand = " ".join([phyml,"--input", tempAln1.name,"--model", model, "-o", "n", "-u", tempAln0.name + "_phyml_tree.txt", ">>", log])
os.system(phymlCommand)
#retrieve
try:
with open(tempAln1.name + "_phyml_stats.txt", "r") as statsFile:
stats = statsFile.read().split()
lnL1 = float(stats[stats.index("Log-likelihood:")+1])
except: lnL1 = np.NaN
#second way validation
#tree
phymlCommand = " ".join([phyml,"--input", tempAln1.name,"--model", model, "-o", opt, ">>", log])
os.system(phymlCommand)
#validation
phymlCommand = " ".join([phyml,"--input", tempAln0.name,"--model", model, "-o", "n", "-u", tempAln1.name + "_phyml_tree.txt", ">>", log])
os.system(phymlCommand)
#retrieve
try:
with open(tempAln0.name + "_phyml_stats.txt", "r") as statsFile:
stats = statsFile.read().split()
lnL0 = float(stats[stats.index("Log-likelihood:")+1])
except: lnL0 = np.NaN
#remove files
if not test:
for f in [f for f in os.listdir(tmpDir) if localName0 in f or localName1 in f]:
os.remove(tmpDir + "/" + f)
return str(lnL0+lnL1)
def raxTree(seqArray,
seqNames,
model,
raxml,
outgroup=None,
prefix="",
test=False,
log="/dev/null"):
#temp file
tempAln = tempfile.NamedTemporaryFile(mode="w",
prefix=prefix,
suffix=".phy",
dir=".",
delete=False)
localName = tempAln.name.rsplit("/", 1)[1]
with tempAln as tA:
tA.write(genomics.makeAlnString(seqNames, seqArray))
if outgroup is not None:
og = " -o " + ",".join(outgroup)
else:
og = ""
#raxCommand = raxml + " -s temp." + uniqueTag + ".phy -n " + uniqueTag + " -m " + model + og + " -V -f d -p 12345 --silent"
raxCommand = raxml + " -s " + tempAln.name + " -n " + localName + " -m " + model + og + " -V -f d -p 12345 --silent >>" + log
if test: print >> sys.stderr, "raxml command:\n", raxCommand
os.system(raxCommand)
#try retrieve the result
try:
treeFile = open("RAxML_bestTree." + localName, "r")
tree = treeFile.readline()
treeFile.close()
except:
tree = "NA\n"
#remove files
if not test:
for f in [f for f in os.listdir(".") if localName in f]:
os.remove(f)
return tree
sys.stderr.write("Getting region " + region + " from geno file...\n")
genoStream = subprocess.Popen(['tabix', '-h', args.genoFile, region],
stdout=subprocess.PIPE)
window = genomics.parseGenoFile(genoStream.stdout,
names=args.samples,
includePositions=True,
splitPhased=args.split)
seqDict = window.seqDict()
seqNames = seqDict.keys()
sys.stderr.write("Extracting CDS...\n")
CDSseqs = [
genomics.CDS(seqDict[name], window.positions,
geneData[scaffold][mRNA]['cdsStarts'],
geneData[scaffold][mRNA]['cdsEnds'],
geneData[scaffold][mRNA]['strand'])
for name in seqNames
]
outputNames = [
name + "_" + mRNA + " " + scaffold + " " +
str(geneData[scaffold][mRNA]['start']) + "-" +
str(geneData[scaffold][mRNA]['end']) for name in seqNames
]
fastaString = genomics.makeAlnString(outputNames,
CDSseqs,
outFormat="fasta",
lineLen=None)
outFile.write(fastaString + "\n")
outFile.close()
if args.regionsFile:
with open(args.regionsFile, "r") as rf:
for line in rf: regions.append(parseRegionList(line.split()))
#only filter and chop sequences if necessary
if len(regions) >= 1:
outNames = []
outSeqs = []
for seqName,start,end,ori in regions:
i = names.index(seqName)
outNames.append(seqName)
if start != None or end != None or ori == "-":
seqLen = len(seqs[i])
if start == None: start = 1
if end == None: end = seqLen
start = max(1,start-args.extendLeft)
end = min(seqLen,end+args.extendRight)
if ori == "-": outSeqs.append(genomics.revTrans(seqs[i][start-1:end]))
else: outSeqs.append(seqs[i][start-1:end])
if not args.preserveNames: outNames[-1] = outNames[-1] + ":" + str(start) + "-" + str(end) + ":" + ori
else: outSeqs.append(seqs[i])
else:
outNames = names
outSeqs = seqs
sys.stderr.write("\nWriting %i sequences.\n" %len(outNames))
sys.stdout.write(genomics.makeAlnString(names=outNames,seqs=outSeqs,outFormat=outFormat,lineLen=l))
#now update if there is data for this window
for siteData in reader.siteBySite(
asDict=True if args.samples else False):
GTs = [siteData["GTs"][name] for name in args.samples
] if args.samples else siteData["GTs"]
siteGTdict[siteData["position"]] = [
genomics.complement(gt) for gt in GTs
] if strand == "-" else GTs
cdsPositions = genomics.CDSpositions(cdsStarts, cdsEnds, strand)
cdsSeqs = [[siteGTdict[position][i] for position in cdsPositions]
for i in range(nSeqs)]
if args.outFormat == "fasta":
outputNames = [
name + "_" + mRNA + " " + scaffold + " " +
str(geneData[scaffold][mRNA]['start']) + "-" +
str(geneData[scaffold][mRNA]['end']) for name in reader.names
]
else:
outputNames = [name + "_" + mRNA for name in reader.names]
alnString = genomics.makeAlnString(outputNames,
cdsSeqs,
outFormat=args.outFormat,
lineLen=None)
outFile.write(alnString + "\n")
outFile.close()
if args.phylipIn: inFormat = "phylip"
else: inFormat = "fasta"
if args.phylipOut: outFormat = "phylip"
else: outFormat = "fasta"
l = args.lineLen
########################################################################
allText = sys.stdin.read()
if inFormat == "fasta": names, seqs = genomics.parseFasta(allText)
else: names, seqs = genomics.parsePhylip(allText)
if args.truncateNames: names = [name.split()[0] for name in names]
if args.seqNames:
outNames = args.seqNames.split(",")
keep = [names.index(name) for name in outNames]
names = [names[k] for k in keep]
seqs = [seqs[k] for k in keep]
if args.start or args.end: seqs = [s[args.start - 1:args.end] for s in seqs]
sys.stdout.write(
genomics.makeAlnString(names=names,
seqs=seqs,
outFormat=outFormat,
lineLen=l))
#############################
samples = args.samples.split(",") if args.samples else None
# if cating all contigs, just parse file and write
if args.mode == "cat":
#read file into window like object
window = genomics.parseGenoFile(genoFile,
names=samples,
splitPhased=args.splitPhased)
#write
seqDict = window.seqDict()
seqFile.write(
genomics.makeAlnString(window.names,
[seqDict[name] for name in window.names],
outFormat=args.format))
genoFile.close()
seqFile.close()
exit()
if args.mode == "windows" or args.mode == "contigs":
if args.mode == "windows":
windType = args.windType
windSize = args.windSize
minSites = args.minSites
stepSize = args.stepSize
overlap = args.overlap
maxDist = args.maxDist
else:
#to get contigs, we just use very lare non-overlapping coordinate windows
