def getHalTreeMutations(halPath, args, rootName=None):
root = rootName
if root is None:
root = getHalRootName(halPath)
for child in getHalChildrenNames(halPath, root):
getHalBranchMutations(halPath, child, args)
getHalTreeMutations(halPath, args, child)
def getHalTreeMutations(halPath, args, rootName=None):
root = rootName
if root is None:
root = getHalRootName(halPath)
for child in getHalChildrenNames(halPath, root):
getHalBranchMutations(halPath, child, args)
getHalTreeMutations(halPath, args, child)
def main(argv=None):
if argv is None:
argv = sys.argv
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter,
description="Compute PhyloP scores (in wig format) for each genome in"
" an alignment. Scores are computed once per column with "
"halPhyloPMP.py and iteratively lifted down the tree using "
"halWiggleLiftover (starting at the root).")
parser.add_argument("hal", help="input hal")
parser.add_argument("mod", help="model file for PhyloP. Can be created "
"with halPhyloPTrain.py")
parser.add_argument("outWigDir", help="directory where output wig files"
" will be written")
parser.add_argument("--root", help="Name of root. If not specified the"
" HAL root will be used", default=None)
parser.add_argument("--numProc",
help="Maximum number of processes.",
type=int, default=1)
parser.add_argument("--bigWig",
help="Run wigToBigWig on each generated wiggle",
action="store_true", default=False)
parser.add_argument("--subtree",
help="Run clade-specific acceleration/conservation on subtree below this node",
default=None)
parser.add_argument("--prec",
help="Number of decimal places in wig output", type=int,
default=None)
# need phyloP options here:
args = parser.parse_args()
if not os.path.isfile(args.hal):
raise RuntimeError("Input hal file %s not found" % args.hal)
if not os.path.isfile(args.mod):
raise RuntimeError("Input mod file %s not found" % args.mod)
if not os.path.isdir(args.outWigDir):
os.makedirs(args.outWigDir)
if not os.path.isdir(args.outWigDir):
raise RuntimeError("%s not found" % args.outWigDir)
args.halGenomes = getHalGenomes(args.hal)
if args.root is None:
args.root = getHalRootName(args.hal)
if not args.root in args.halGenomes:
raise RuntimeError("Root genome %s not found." % args.root)
if args.subtree is not None and args.root not in args.halGenomes:
raise RuntimeError("Subtree root %s not found." % args.subtree)
# make a little id tag for temporary maf slices
S = string.ascii_uppercase + string.digits
args.tempID = 'halTreePhyloP' + ''.join(random.choice(S) for x in range(5))
computeTreePhyloP(args)
def runParallelSlices(options):
refGenome = options.refGenome
if refGenome is None:
refGenome = getHalRootName(options.halFile)
refSequenceStats = getHalSequenceStats(options.halFile, refGenome)
options.smallFile = False
options.firstSmallFile = True
sliceCmds = []
sliceOpts = []
# we are going to deal with sequence coordinates
if options.splitBySequence is True or options.refSequence is not None:
for sequence, seqLen, nt, nb in refSequenceStats:
if options.refSequence is None or sequence == options.refSequence:
seqOpts = copy.deepcopy(options)
if seqLen < options.smallSize:
seqOpts.smallFile = True
seqOpts.refGenome = refGenome
seqOpts.refSequence = sequence
index = 0
for sStart, sLen, sIdx in computeSlices(seqOpts, seqLen):
seqOpts.start = sStart
seqOpts.length = sLen
seqOpts.sliceNumber = sIdx
sliceCmds.append(getHal2MafCmd(seqOpts))
sliceOpts.append(copy.deepcopy(seqOpts))
if seqOpts.smallFile is True and seqLen > 0:
options.firstSmallFile = False
# we are slicing the gnome coordinates directly
else:
seqOpts = copy.deepcopy(options)
assert seqOpts.splitBySequence is False
genomeLen = getHalGenomeLength(seqOpts.halFile, refGenome)
# auto compute slice size from numprocs
if seqOpts.sliceSize == None and seqOpts.numProc > 1:
refLen = genomeLen
if seqOpts.length is not None and seqOpts.length > 0:
refLen = seqOpts.length
seqOpts.sliceSize = int(math.ceil(refLen / seqOpts.numProc))
index = 0
for sStart, sLen, sIdx in computeSlices(seqOpts, genomeLen):
seqOpts.start = sStart
seqOpts.length = sLen
seqOpts.sliceNumber = sIdx
sliceCmds.append(getHal2MafCmd(seqOpts))
sliceOpts.append(copy.deepcopy(seqOpts))
# run in parallel
runParallelShellCommands(sliceCmds, options.numProc)
# concatenate into output if desired
concatenateSlices(sliceOpts, sliceCmds)
def runParallelSlices(options):
refGenome = options.refGenome
if refGenome is None:
refGenome = getHalRootName(options.halFile)
refSequenceStats = getHalSequenceStats(options.halFile, refGenome)
options.smallFile = False
options.firstSmallFile = True
sliceCmds = []
sliceOpts = []
# we are going to deal with sequence coordinates
if options.splitBySequence is True or options.refSequence is not None:
for sequence, seqLen, nt, nb in refSequenceStats:
if options.refSequence is None or sequence == options.refSequence:
seqOpts = copy.deepcopy(options)
if seqLen < options.smallSize:
seqOpts.smallFile = True
seqOpts.refGenome = refGenome
seqOpts.refSequence = sequence
index = 0
for sStart, sLen, sIdx in computeSlices(seqOpts, seqLen):
seqOpts.start = sStart
seqOpts.length = sLen
seqOpts.sliceNumber = sIdx
sliceCmds.append(getHal2MafCmd(seqOpts))
sliceOpts.append(copy.deepcopy(seqOpts))
if seqOpts.smallFile is True and seqLen > 0:
options.firstSmallFile = False
# we are slicing the gnome coordinates directly
else:
seqOpts = copy.deepcopy(options)
assert seqOpts.splitBySequence is False
genomeLen = getHalGenomeLength(seqOpts.halFile, refGenome)
# auto compute slice size from numprocs
if seqOpts.sliceSize == None and seqOpts.numProc > 1:
refLen = genomeLen
if seqOpts.length is not None and seqOpts.length > 0:
refLen = seqOpts.length
seqOpts.sliceSize = int(math.ceil(refLen / seqOpts.numProc))
index = 0
for sStart, sLen, sIdx in computeSlices(seqOpts, genomeLen):
seqOpts.start = sStart
seqOpts.length = sLen
seqOpts.sliceNumber = sIdx
sliceCmds.append(getHal2MafCmd(seqOpts))
sliceOpts.append(copy.deepcopy(seqOpts))
# run in parallel
runParallelShellCommands(sliceCmds, options.numProc)
# concatenate into output if desired
concatenateSlices(sliceOpts, sliceCmds)
def runParallelSlices(options):
refGenome = options.refGenome
if refGenome is None:
refGenome = getHalRootName(options.halFile)
refSequenceStats = getHalSequenceStats(options.halFile, refGenome)
options.smallFile = False
options.firstSmallFile = True
sliceCmds = []
sliceOpts = []
if options.refSequence is not None:
refStat = [x for x in refSequenceStats if x[1] ==
options.refSequence]
if len(refStat != 1):
raise RuntimeError("Sequence %s not found in genome %s" % (
options.refSequence, options.refGenome))
totalLength = int(refStat[1])
else:
totalLength = getHalGenomeLength(options.halFile, refGenome)
seqOpts = copy.deepcopy(options)
# auto compute slice size from numprocs
if seqOpts.sliceSize == None and seqOpts.numProc > 1:
refLen = totalLength
if seqOpts.length is not None and seqOpts.length > 0:
refLen = seqOpts.length
seqOpts.sliceSize = int(math.ceil(refLen / seqOpts.numProc))
index = 0
for sStart, sLen, sIdx in computeSlices(seqOpts, totalLength):
seqOpts.start = sStart
seqOpts.length = sLen
seqOpts.sliceNumber = sIdx
sliceCmds.append(getHalPhyloPCmd(seqOpts))
sliceOpts.append(copy.deepcopy(seqOpts))
# run in parallel
runParallelShellCommands(sliceCmds, options.numProc)
# concatenate into output if desired
concatenateSlices(sliceOpts, sliceCmds)
writeChromSizes(options)
def runParallelSlices(options):
refGenome = options.refGenome
if refGenome is None:
refGenome = getHalRootName(options.halFile)
refSequenceStats = getHalSequenceStats(options.halFile, refGenome)
options.smallFile = False
options.firstSmallFile = True
sliceCmds = []
sliceOpts = []
if options.refSequence is not None:
refStat = [x for x in refSequenceStats if x[1] == options.refSequence]
if len(refStat != 1):
raise RuntimeError("Sequence %s not found in genome %s" %
(options.refSequence, options.refGenome))
totalLength = int(refStat[1])
else:
totalLength = getHalGenomeLength(options.halFile, refGenome)
seqOpts = copy.deepcopy(options)
# auto compute slice size from numprocs
if seqOpts.sliceSize == None and seqOpts.numProc > 1:
refLen = totalLength
if seqOpts.length is not None and seqOpts.length > 0:
refLen = seqOpts.length
seqOpts.sliceSize = int(math.ceil(refLen / seqOpts.numProc))
index = 0
for sStart, sLen, sIdx in computeSlices(seqOpts, totalLength):
seqOpts.start = sStart
seqOpts.length = sLen
seqOpts.sliceNumber = sIdx
sliceCmds.append(getHalPhyloPCmd(seqOpts))
sliceOpts.append(copy.deepcopy(seqOpts))
# run in parallel
runParallelShellCommands(sliceCmds, options.numProc)
# concatenate into output if desired
concatenateSlices(sliceOpts, sliceCmds)
writeChromSizes(options)
def runParallelSlices(options):
refGenome = options.refGenome
if refGenome is None:
refGenome = getHalRootName(options.halFile)
options.smallFile = False
options.firstSmallFile = True
if options.refTargets:
sliceCmds, sliceOpts = partitionRefTargets(options)
elif options.splitBySequence is True or options.refSequence is not None:
sliceCmds, sliceOpts = partitionBySeqCoords(options, refGenome)
else:
sliceCmds, sliceOpts = partitionByGenomeCoords(options, refGenome)
# run in parallel
runParallelShellCommands(sliceCmds, options.numProc)
# clean up temporary bed files (if present)
for opts in sliceOpts:
if opts.refTargets and os.path.isfile(opts.refTargets):
os.remove(opts.refTargets)
# concatenate into output if desired
concatenateSlices(sliceOpts, sliceCmds)
