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))
class ProjectWrapper:
alignmentDirName = 'progressiveAlignment'
def __init__(self, options, seqFile, workingDir):
self.options = options
self.seqFile = seqFile
self.workingDir = workingDir
self.configWrapper = None
self.expWrapper = None
self.processConfig()
self.processExperiment()
def processConfig(self):
# read in the default right out of cactus
if self.options.configFile is not None:
configPath = self.options.configFile
else:
dir = os.path.join(cactusRootPath(), "progressive")
configPath = os.path.join(dir,
"cactus_progressive_workflow_config.xml")
configXml = ET.parse(configPath).getroot()
self.configWrapper = ConfigWrapper(configXml)
# here we can go through the options and apply some to the config
self.configWrapper.setBuildHal(True)
self.configWrapper.setBuildFasta(True)
if self.options.outputMaf is not None:
self.configWrapper.setBuildMaf(True)
self.configWrapper.setJoinMaf(True)
# pre-emptively turn down maxParallelSubtree for singleMachine
# mode if not enough threads are provided to support it. Probably
# need to do something for other ?combined? batch systems?
if self.options.batchSystem == 'singleMachine' and \
self.options.database == 'kyoto_tycoon':
if int(self.options.maxThreads) < \
self.configWrapper.getMaxParallelSubtrees() * 3:
self.configWrapper.setMaxParallelSubtrees(
max(1, int(self.options.maxThreads) / 3))
# this is a little hack to effectively toggle back to the
# non-progressive version of cactus (as published in Gen. Res. 2011)
# from the high-level interface.
if self.options.legacy is True:
self.configWrapper.setSubtreeSize(sys.maxint)
def processExperiment(self):
expXml = self.seqFile.toXMLElement()
#create the cactus disk
cdElem = ET.SubElement(expXml, "cactus_disk")
database = self.options.database
assert database == "kyoto_tycoon" or database == "tokyo_cabinet"
confElem = ET.SubElement(cdElem, "st_kv_database_conf")
confElem.attrib["type"] = database
dbElem = ET.SubElement(confElem, database)
self.expWrapper = ExperimentWrapper(expXml)
if self.options.database == "kyoto_tycoon":
self.expWrapper.setDbPort(str(self.options.ktPort))
if self.options.ktType == 'memory':
self.expWrapper.setDbInMemory(True)
self.expWrapper.setDbSnapshot(False)
elif self.options.ktType == 'snapshot':
self.expWrapper.setDbInMemory(True)
self.expWrapper.setDbSnapshot(True)
else:
assert self.options.ktType == 'disk'
self.expWrapper.setDbInMemory(False)
self.expWrapper.setDbSnapshot(False)
# sonlib doesn't allow for spaces in attributes in the db conf
# which renders this options useless
# if self.options.ktOpts is not None:
# self.expWrapper.setDbServerOptions(self.options.ktOpts)
if self.options.ktCreateTuning is not None:
self.expWrapper.setDbCreateTuningOptions(
self.options.ktCreateTuning)
if self.options.ktOpenTuning is not None:
self.expWrapper.setDbReadTuningOptions(
self.options.ktOpenTuning)
def writeXml(self):
assert os.path.isdir(self.workingDir)
configPath = os.path.abspath(
os.path.join(self.workingDir, "config.xml"))
expPath = os.path.abspath(
os.path.join(self.workingDir, "expTemplate.xml"))
self.expWrapper.setConfigPath(configPath)
self.configWrapper.writeXML(configPath)
self.expWrapper.writeXML(expPath)
projPath = os.path.join(self.workingDir,
ProjectWrapper.alignmentDirName)
if os.path.exists(projPath) and self.options.overwrite:
system("rm -rf %s" % projPath)
if self.options.outputMaf is True:
fixNames=1
else:
fixNames=0
if os.path.exists(projPath):
if not self.isSameAsExisting(expPath, projPath, fixNames):
raise RuntimeError("Existing project %s not " % projPath+
"compatible with current input. Please "
"erase the working directory or rerun "
"with the --overwrite option to start "
"from scratch.")
else:
logPath = os.path.join(self.workingDir, 'cactus.log')
logFile = open(logPath, "a")
logFile.write("\nContinuing existing alignment. Use "
"--overwrite or erase the working directory to "
"force restart from scratch.\n")
logFile.close()
else:
cmd = "cactus_createMultiCactusProject.py %s %s --fixNames=%d" % (
expPath, projPath, fixNames)
if len(self.seqFile.outgroups) > 0:
cmd += " --outgroupNames " + ",".join(self.seqFile.outgroups)
system(cmd)
# create a project in a dummy directory. check if the
# project xml is the same as the current project.
# we do this to see if we should start fresh or try to
# work with the existing project when the overwrite flag is off
def isSameAsExisting(self, expPath, projPath, fixNames):
if not os.path.exists(projPath):
return False
oldPath = os.path.dirname(projPath + "/")
tempPath = "%s_temp" % oldPath
if os.path.exists(tempPath):
system("rm -rf %s" % tempPath)
cmd = "cactus_createMultiCactusProject.py %s %s --fixNames=%d" % (
expPath, tempPath, fixNames)
if len(self.seqFile.outgroups) > 0:
cmd += " --outgroupNames " + ",".join(self.seqFile.outgroups)
system(cmd)
projFilePathNew = os.path.join(tempPath,'%s_temp_project.xml' %
self.alignmentDirName)
projFilePathOld = os.path.join(oldPath, '%s_project.xml' %
self.alignmentDirName)
newFile = [line for line in open(projFilePathNew, "r")]
oldFile = [line for line in open(projFilePathOld, "r")]
areSame = True
if len(newFile) != len(oldFile):
areSame = False
for newLine, oldLine in zip(newFile, oldFile):
if newLine.replace(tempPath, oldPath) != oldLine:
areSame = False
system("rm -rf %s" % tempPath)
return areSame