def main():
parser = ArgumentParser(description=__doc__)
parser.add_argument('hal', help='hal file')
parser.add_argument('refGenome', help='reference genome')
parser.add_argument('halTreeMutationsDir',
help='the directory output by halTreeMutations.py')
parser.add_argument(
'--targets',
help='target genomes (comma-separated), default: all leaves')
parser.add_argument('outputDir',
help='output directory for reference beds')
opts = parser.parse_args()
# Get the species tree from the hal file.
newickTree = popenCatch('halStats --tree %s' % (opts.hal))
tree = NXNewick().parseString(newickTree)
# Set the target genomes to be all leaves (minus the reference) if not otherwise directed.
leafGenomes = [tree.getName(x) for x in tree.getLeaves()]
if opts.refGenome not in leafGenomes:
raise ValueError("Reference genome %s is not a leaf genome." %
opts.refGenome)
if opts.targets is None:
opts.targets = [x for x in leafGenomes if x != opts.refGenome]
else:
opts.targets = opts.targets.split(',')
if not all([x in leafGenomes for x in opts.targets]):
raise ValueError("Some target genomes are not leaves.")
try:
os.makedirs(opts.outputDir)
except:
if not os.path.isdir(opts.outputDir):
raise
for target in opts.targets:
refID = getTreeID(tree, opts.refGenome)
targetID = getTreeID(tree, target)
mrca = getMRCA(tree, refID, targetID)
pathToTarget = getPath(opts.hal, opts.refGenome, target)
pathUp, pathDown = [
list(v) for k, v in groupby(
pathToTarget, lambda x: x == tree.getName(mrca)) if k != True
]
bedForTarget = os.path.join(opts.outputDir, target + '.bed')
# First, walk up the tree to the MRCA.
for curGenome in pathUp:
liftMutations(opts.halTreeMutationsDir,
opts.hal,
curGenome,
opts.refGenome,
bedForTarget,
reversePolarity=True)
# Next, walk down the tree to the target.
for curGenome in pathDown:
liftMutations(opts.halTreeMutationsDir, opts.hal, curGenome,
opts.refGenome, bedForTarget)
def getCactusWorkflowExperimentForTest(sequences,
newickTreeString,
outputDir,
configFile=None,
constraints=None,
progressive=False,
reconstruct=True):
"""Wrapper to constructor of CactusWorkflowExperiment which additionally incorporates
any globally set database conf.
"""
halFile = os.path.join(outputDir, "test.hal")
fastaFile = os.path.join(outputDir, "test.fa")
databaseConf = ET.fromstring(
_GLOBAL_DATABASE_CONF_STRING
) if _GLOBAL_DATABASE_CONF_STRING is not None else None
tree = NXNewick().parseString(newickTreeString, addImpliedRoots=False)
genomes = [
tree.getName(id) for id in tree.postOrderTraversal() if tree.isLeaf(id)
]
exp = ExperimentWrapper.createExperimentWrapper(newickTreeString,
genomes,
outputDir,
databaseConf=databaseConf,
configFile=configFile,
halFile=halFile,
fastaFile=fastaFile,
constraints=constraints,
progressive=progressive)
for genome, sequence in zip(genomes, sequences):
print((genome, sequence))
exp.setSequenceID(genome, sequence)
exp.setRootGenome("reference")
if reconstruct:
exp.setRootReconstructed(True)
return exp
def run(self):
# Find all ancestral genomes using the tree.
newickStr = popenCatch("halStats --tree %s" % self.halFile)
tree = NXNewick().parseString(newickStr)
bedFiles = {} # genome => bed files of inserted columns
for nodeId in tree.postOrderTraversal():
if len(tree.getChildren(nodeId)) == 0:
# leaf node, skip
continue
assert tree.hasName(nodeId)
genome = tree.getName(nodeId)
bedFileForGenome = getTempFile(rootDir=self.getGlobalTempDir())
bedFiles[genome] = bedFileForGenome
self.addChildTarget(GetInsertedColumnBed(self.halFile, genome, bedFileForGenome))
self.setFollowOnTarget(RunAncestorsMLParallel(self.halFile, self.phyloPModel, bedFiles, self.jobsPerGenome, self.threshold))
if __name__ == '__main__':
# Parse args
if len(sys.argv) < 3:
print __doc__
sys.exit(1)
newickPath = sys.argv[1]
fastaPath = sys.argv[2]
treeString = open(newickPath).read().split("\n")[0].strip()
tree = NXNewick().parseString(treeString)
sequences = {}
for name, seq in fastaRead(open(fastaPath)):
sequences[name] = seq
# Print MAF, with sequence lines in post-order.
print '##maf version=1 scoring=NA'
print 'a tree="%s"' % (treeString)
for nodeId in tree.postOrderTraversal():
if not tree.isLeaf(nodeId):
continue
nodeName = tree.getName(nodeId)
if nodeName not in sequences:
raise RuntimeError("The tree has a node %s which was not found in the fasta file" % (nodeName))
seq = sequences[nodeName]
seqLen = lengthWithoutGaps(seq)
print 's %s 0 %d + %d %s' % (nodeName, seqLen, seqLen, seq)
# mafValidator wants an empty closing line(?)
print ''
