def runProgressive(self):
logger.debug("Going to put the alignment in %s" % self.outputDir)
if not os.path.isdir(self.outputDir):
os.mkdir(self.outputDir)
if not os.path.exists(os.path.join(self.outputDir, "progressiveCactusAlignment")):
xmlTree = ET.parse(os.path.join(getRootPathString(), "lib", "cactus_workflow_config.xml"))
#Set the parameters
tempLocalDir = os.path.join(self.outputDir, "tempProgressiveCactusAlignment")
system("rm -rf %s" % tempLocalDir)
os.mkdir(tempLocalDir)
#Set the config parameters
self.params.applyToXml(xmlTree)
config = xmlTree.getroot()
assert config is not None
#Write the config file
tempConfigFile = os.path.join(tempLocalDir, "config.xml")
fileHandle = open(tempConfigFile, 'w')
assert fileHandle is not None
tree = ET.ElementTree(config)
tree.write(fileHandle)
fileHandle.close()
#Make the experiment file
tempExperimentFile = os.path.join(tempLocalDir, "experiment.xml")
if self.params.kyotoTycoon == True:
dbConfElem = ET.Element("st_kv_database_conf", type="kyoto_tycoon")
ktElem = ET.SubElement(dbConfElem, "kyoto_tycoon", host="localhost", port="1978", database_dir="dummy")
else:
dbConfElem = None
cactusWorkflowExperiment = CactusWorkflowExperiment(
sequences=self.sequences,
newickTreeString=self.newickTree,
#requiredSpecies=self.requiredSpecies,
#singleCopySpecies=self.singleCopySpecies,
databaseName="cactusAlignment",
outputDir=tempLocalDir,
configFile=tempConfigFile,
databaseConf = dbConfElem)
cactusWorkflowExperiment.writeExperimentFile(tempExperimentFile)
#The jobtree
tempJobTreeDir = os.path.join(tempLocalDir, "jobTree")
#The place to put the temporary experiment dir
tempExperimentDir = os.path.join(tempLocalDir, "progressiveCactusAlignment")
#The temporary experiment
runCactusCreateMultiCactusProject(tempExperimentFile,
tempExperimentDir)
logger.info("Setup the cactus progressive experiment")
runCactusProgressive(os.path.join(tempExperimentDir, "progressiveCactusAlignment_project.xml"),
tempJobTreeDir,
#batchSystem=batchSystem,
buildMaf=True,
joinMaf=True,
#buildTrees=buildTrees, buildFaces=buildFaces, buildReference=buildReference,
jobTreeStats=True,
maxThreads=4,
logLevel="DEBUG")
logger.info("Ran the progressive workflow")
#Check if the jobtree completed sucessively.
runJobTreeStatusAndFailIfNotComplete(tempJobTreeDir)
logger.info("Checked the job tree dir for the progressive run")
#Run the cactus tree stats
expPath = os.path.join(tempExperimentDir, "Anc0", "Anc0_experiment.xml")
exp = ExperimentWrapper(ET.parse(expPath).getroot())
if exp.getDbType() == "kyoto_tycoon":
ktserver = KtserverLauncher()
ktserver.spawnServer(exp)
treeStatsFile = os.path.join(self.outputDir, "treeStats.xml")
system("cactus_treeStats --cactusDisk \'%s\' --flowerName 0 --outputFile %s" %(exp.getDiskDatabaseString(),
treeStatsFile))
if exp.getDbType() == "kyoto_tycoon":
ktserver.killServer(exp)
#Now copy the true assembly back to the output
system("mv %s %s/experiment.xml" % (tempExperimentFile, self.outputDir))
system("mv %s %s" % (tempExperimentDir, self.outputDir))
system("jobTreeStats --jobTree %s --outputFile %s/jobTreeStats.xml" % (tempJobTreeDir, self.outputDir))
system("mv %s %s/config.xml" % (tempConfigFile, self.outputDir))
#But keep a link to the multicactus project in its original path so we can navigate
# the paths in the xml...
actualResultsDir = os.path.join(os.path.abspath(self.outputDir), "progressiveCactusAlignment")
tempResultsDir = os.path.join(self.outputDir, "tempProgressiveCactusAlignment")
system("ln -s %s %s" % (actualResultsDir, tempResultsDir))
def run(self):
if not os.path.isdir(self.outputDir):
os.mkdir(self.outputDir)
cactusAlignmentName = "cactusAlignment"
outputFile = os.path.join(self.outputDir, cactusAlignmentName)
if not os.path.exists(outputFile):
config = ET.parse(
os.path.join(getRootPathString(), "lib",
"cactus_workflow_config.xml")).getroot()
#Set the reference algorithm
config.find("reference").attrib[
"matching_algorithm"] = self.referenceAlgorithm
#Do the minimum block degree configuration
iterations = config.find("alignment").find("iterations")
blastIteration = iterations.findall("iteration")[0]
baseIteration = iterations.findall("iteration")[1]
minimumBlastBlockDegree = self.minimumBlockDegree
if minimumBlastBlockDegree <= 1:
minimumBlastBlockDegree = 2
blastIteration.find("core").attrib["minimumBlockDegree"] = str(
minimumBlastBlockDegree)
baseIteration.attrib["minimumBlockDegree"] = str(
self.minimumBlockDegree)
baseIteration.attrib["prune_out_stub_alignments"] = str(
int(self.pruneOutStubAlignments))
baseIteration.attrib["gap_gamma"] = str(float(self.gapGamma))
#Set the blast string
blastIteration.find(
"blast").attrib["blastString"] = blastIteration.find(
"blast").attrib["blastString"].replace(
"PARAMETERS", self.blastAlignmentString)
blastIteration.find(
"blast").attrib["selfBlastString"] = blastIteration.find(
"blast").attrib["selfBlastString"].replace(
"PARAMETERS", self.blastAlignmentString)
#Get rid of the base level, if needed
if not self.baseLevel:
iterations.remove(baseIteration)
#Set the number of chains to allow in a level, during promotion
config.find("normal").attrib["max_number_of_chains"] = str(
self.maxNumberOfChains)
#Set the number of chains to order per round of the matching algorithm
config.find("reference").attrib["permutations"] = str(
self.permutations)
#Set the chain weight function
if bool(self.useSimulatedAnnealing):
config.find("reference").attrib["useSimulatedAnnealing"] = "1"
config.find("reference").attrib["theta"] = str(self.theta)
#Write the config file
tempConfigFile = os.path.join(self.getLocalTempDir(), "config.xml")
fileHandle = open(tempConfigFile, 'w')
tree = ET.ElementTree(config)
tree.write(fileHandle)
fileHandle.close()
#Make the supporting temporary files
tempExperimentFile = os.path.join(self.getLocalTempDir(),
"experiment.xml")
tempJobTreeDir = os.path.join(self.getLocalTempDir(), "jobTree")
#Make the experiment file
cactusWorkflowExperiment = CactusWorkflowExperiment(
sequences=self.sequences.split(),
newickTreeString=self.options.newickTree,
requiredSpecies=[(1, self.requiredSpecies.split())],
singleCopySpecies=self.singleCopySpecies,
outgroupEvent=self.options.outgroupEvent,
databaseName=cactusAlignmentName,
outputDir=self.getLocalTempDir(),
configFile=tempConfigFile)
cactusWorkflowExperiment.writeExperimentFile(tempExperimentFile)
#Now run cactus workflow
runCactusWorkflow(experimentFile=tempExperimentFile,
jobTreeDir=tempJobTreeDir,
setupAndBuildAlignments=True,
buildTrees=False,
buildFaces=False,
buildReference=True,
batchSystem="single_machine",
maxThreads=1,
jobTreeStats=True)
logger.info("Ran the workflow")
#Check if the jobtree completed sucessively.
runJobTreeStatusAndFailIfNotComplete(tempJobTreeDir)
logger.info("Checked the job tree dir")
#Now copy the true assembly back to the output
system("mv %s %s/experiment.xml" %
(tempExperimentFile, self.outputDir))
system("mv %s %s/config.xml" % (tempConfigFile, self.outputDir))
#Copy across the final alignment
localCactusDisk = os.path.join(self.getLocalTempDir(),
cactusAlignmentName)
#Move the final db
system("mv %s %s" % (localCactusDisk, outputFile))
#Compute the stats
system(
"jobTreeStats --jobTree %s --outputFile %s/jobTreeStats.xml" %
(tempJobTreeDir, self.outputDir))
#We're done!
self.addChildTarget(MakeStats(outputFile, self.outputDir,
self.options))
def runVanilla(self):
logger.debug("Going to put the alignment in %s" % self.outputDir)
if not os.path.isdir(self.outputDir):
os.mkdir(self.outputDir)
if not os.path.exists(os.path.join(self.outputDir, "cactusAlignmentVanilla")):
xmlTree = ET.parse(os.path.join(getRootPathString(), "lib", "cactus_workflow_config.xml"))
#Set the parameters
tempLocalDir = os.path.join(self.outputDir, "tempVanillaCactusAlignment")
system("rm -rf %s" % tempLocalDir)
os.mkdir(tempLocalDir)
#Set the config parameters
self.params.applyToXml(xmlTree)
config = xmlTree.getroot()
assert config is not None
#Write the config file
tempConfigFile = os.path.join(tempLocalDir, "config.xml")
fileHandle = open(tempConfigFile, 'w')
assert fileHandle is not None
tree = ET.ElementTree(config)
tree.write(fileHandle)
fileHandle.close()
#Make the experiment file
tempExperimentFile = os.path.join(tempLocalDir, "experiment.xml")
#Now do standard cactus..
#Make the experiment file
tempExperimentFile2 = os.path.join(tempLocalDir, "experiment.xml")
cactusWorkflowExperiment = CactusWorkflowExperiment(
sequences=self.sequences,
newickTreeString=self.newickTree,
#requiredSpecies=self.requiredSpecies,
#singleCopySpecies=self.singleCopySpecies,
databaseName="cactusAlignmentVanilla",
outputDir=tempLocalDir,
configFile=tempConfigFile)
tempExperimentDir2 = os.path.join(tempLocalDir, "cactusAlignmentVanilla")
cactusWorkflowExperiment.writeExperimentFile(tempExperimentFile2)
# apply naming to the event tree to be consistent with progressive
exp = ExperimentWrapper(ET.parse(tempExperimentFile2).getroot())
cleanEventTree(exp)
exp.writeXML(tempExperimentFile2)
#We're done with the progressive, now run the vanilla cactus for comparison
tempJobTreeDir2 = os.path.join(tempLocalDir, "jobTreeVanilla")
runCactusWorkflow(tempExperimentFile2, tempJobTreeDir2,
jobTreeStats=True,
setupAndBuildAlignments=True,
buildReference=True,
maxThreads=4)
runJobTreeStatusAndFailIfNotComplete(tempJobTreeDir2)
logger.info("Checked the job tree dir for the vanilla run")
runCactusMAFGenerator(os.path.join(self.outputDir, "cactusVanilla.maf"), getCactusDiskString(tempExperimentDir2))
#Run the cactus tree stats
treeStatsFile = os.path.join(self.outputDir, "treeStats.xml")
system("cactus_treeStats --cactusDisk \'%s\' --flowerName 0 --outputFile %s" %(exp.getDiskDatabaseString(),
treeStatsFile))
system("jobTreeStats --jobTree %s --outputFile %s/jobTreeStats.xml" % (tempJobTreeDir2, self.outputDir))
system("mv %s %s" % (tempExperimentDir2, self.outputDir))
system("mv %s %s/experiment.xml" % (tempExperimentFile2, self.outputDir))
def run(self):
if not os.path.isdir(self.outputDir):
os.mkdir(self.outputDir)
cactusAlignmentName = "cactusAlignment"
outputFile = os.path.join(self.outputDir, cactusAlignmentName)
if not os.path.exists(outputFile):
config = ET.parse(os.path.join(getRootPathString(), "lib", "cactus_workflow_config.xml")).getroot()
#Set the reference algorithm
config.find("reference").attrib["matching_algorithm"] = self.referenceAlgorithm
#Do the minimum block degree configuration
iterations = config.find("alignment").find("iterations")
blastIteration = iterations.findall("iteration")[0]
baseIteration = iterations.findall("iteration")[1]
minimumBlastBlockDegree = self.minimumBlockDegree
if minimumBlastBlockDegree <= 1:
minimumBlastBlockDegree = 2
blastIteration.find("core").attrib["minimumBlockDegree"] = str(minimumBlastBlockDegree)
baseIteration.attrib["minimumBlockDegree"] = str(self.minimumBlockDegree)
baseIteration.attrib["prune_out_stub_alignments"] = str(int(self.pruneOutStubAlignments))
baseIteration.attrib["gap_gamma"] = str(float(self.gapGamma))
#Set the blast string
blastIteration.find("blast").attrib["blastString"] = blastIteration.find("blast").attrib["blastString"].replace("PARAMETERS", self.blastAlignmentString)
blastIteration.find("blast").attrib["selfBlastString"] = blastIteration.find("blast").attrib["selfBlastString"].replace("PARAMETERS", self.blastAlignmentString)
#Get rid of the base level, if needed
if not self.baseLevel:
iterations.remove(baseIteration)
#Set the number of chains to allow in a level, during promotion
config.find("normal").attrib["max_number_of_chains"] = str(self.maxNumberOfChains)
#Set the number of chains to order per round of the matching algorithm
config.find("reference").attrib["permutations"] = str(self.permutations)
#Set the chain weight function
if bool(self.useSimulatedAnnealing):
config.find("reference").attrib["useSimulatedAnnealing"]="1"
config.find("reference").attrib["theta"] = str(self.theta)
#Write the config file
tempConfigFile = os.path.join(self.getLocalTempDir(), "config.xml")
fileHandle = open(tempConfigFile, 'w')
tree = ET.ElementTree(config)
tree.write(fileHandle)
fileHandle.close()
#Make the supporting temporary files
tempExperimentFile = os.path.join(self.getLocalTempDir(), "experiment.xml")
tempJobTreeDir = os.path.join(self.getLocalTempDir(), "jobTree")
#Make the experiment file
cactusWorkflowExperiment = CactusWorkflowExperiment(
sequences=self.sequences.split(),
newickTreeString=self.options.newickTree,
requiredSpecies=[ (1, self.requiredSpecies.split() ) ],
singleCopySpecies=self.singleCopySpecies,
outgroupEvent = self.options.outgroupEvent,
databaseName=cactusAlignmentName,
outputDir=self.getLocalTempDir(),
configFile=tempConfigFile)
cactusWorkflowExperiment.writeExperimentFile(tempExperimentFile)
#Now run cactus workflow
runCactusWorkflow(experimentFile=tempExperimentFile, jobTreeDir=tempJobTreeDir,
setupAndBuildAlignments=True,
buildTrees=False, buildFaces=False, buildReference=True,
batchSystem="single_machine", maxThreads=1, jobTreeStats=True)
logger.info("Ran the workflow")
#Check if the jobtree completed sucessively.
runJobTreeStatusAndFailIfNotComplete(tempJobTreeDir)
logger.info("Checked the job tree dir")
#Now copy the true assembly back to the output
system("mv %s %s/experiment.xml" % (tempExperimentFile, self.outputDir))
system("mv %s %s/config.xml" % (tempConfigFile, self.outputDir))
#Copy across the final alignment
localCactusDisk = os.path.join(self.getLocalTempDir(), cactusAlignmentName)
#Move the final db
system("mv %s %s" % (localCactusDisk, outputFile))
#Compute the stats
system("jobTreeStats --jobTree %s --outputFile %s/jobTreeStats.xml" % (tempJobTreeDir, self.outputDir))
#We're done!
self.addChildTarget(MakeStats(outputFile, self.outputDir, self.options))