def run(self):
classifiers = classes_in_module(src.classifiers)
attributes = classes_in_module(src.attributes)
classifyDb = os.path.join(self.outDir, "classify.db")
detailsDb = os.path.join(self.outDir, "details.db")
attributesDb = os.path.join(self.outDir, "attributes.db")
self.initializeDb(classifyDb, classifiers, dataType="INTEGER")
self.initializeDb(detailsDb, classifiers, dataType="TEXT")
for classifier, genome in product(classifiers, self.genomes):
classifyDict = pickle.load(
open(
os.path.join(self.tmpDir, genome,
"Classify" + classifier.__name__ + genome),
"rb"))
self.simpleUpdateWrapper(classifyDict, classifyDb, genome,
classifier.__name__)
detailsDict = pickle.load(
open(
os.path.join(self.tmpDir, genome,
"Details" + classifier.__name__ + genome),
"rb"))
self.simpleBedUpdateWrapper(detailsDict, detailsDb, genome,
classifier.__name__)
self.initializeDb(attributesDb, attributes)
for attribute, genome in product(attributes, self.genomes):
attributeDict = pickle.load(
open(
os.path.join(self.tmpDir, genome,
"Attribute" + attribute.__name__ + genome),
"rb"))
self.simpleUpdateWrapper(attributeDict, attributesDb, genome,
attribute.__name__)
def run_aug_classifiers(args, target, tmp_dir):
aug_classifiers = classes_in_module(src.augustus_classifiers)
for classifier in aug_classifiers:
target.addChildTarget(
classifier(args.refFasta, args.annotationGp, args.refGenome,
tmp_dir, args.genome, args.psl, args.refPsl, args.fasta,
args.targetGp, args.augustusGp))
# in Augustus mode we run the alignment-free classifiers on augustus transcripts
ref_classifiers = classes_in_module(src.classifiers)
for classifier in ref_classifiers:
target.addChildTarget(
classifier(args.fasta, args.augustusGp, args.genome, tmp_dir))
def build_analyses(target, ref_genome, genome, annotation_gp, psl, gp, fasta, ref_fasta, sizes, gencode_attributes,
out_dir):
# find all user-defined classes in the categories of analyses
out_file_tree = TempFileTree(target.getGlobalTempDir())
classifiers = classes_in_module(src.classifiers)
attributes = classes_in_module(src.attributes)
for classifier in classifiers:
target.addChildTarget(classifier(genome, psl, fasta, ref_fasta, annotation_gp, gencode_attributes, gp,
ref_genome, out_file_tree))
for attribute in attributes:
target.addChildTarget(attribute(genome, psl, fasta, ref_fasta, annotation_gp, gencode_attributes, gp,
ref_genome, out_file_tree))
# merge the resulting pickled files into sqlite databases and construct BED tracks
target.setFollowOnTargetFn(database, memory=8 * (1024 ** 3),
args=(out_dir, genome, psl, sizes, gp, annotation_gp, out_file_tree))
def run(self):
augustusClassifiers = classes_in_module(src.augustusClassifiers)
classifyDb = os.path.join(self.outDir, "augustusClassify.db")
if os.path.exists(classifyDb):
os.remove(classifyDb)
detailsDb = os.path.join(self.outDir, "augustusDetails.db")
if os.path.exists(detailsDb):
os.remove(detailsDb)
self.initializeDb(classifyDb, augustusClassifiers, dataType="INTEGER")
self.initializeDb(detailsDb, augustusClassifiers, dataType="TEXT")
for classifier, genome in product(augustusClassifiers, self.genomes):
classifyDict = pickle.load(
open(
os.path.join(self.tmpDir, genome,
"Classify" + classifier.__name__ + genome),
"rb"))
self.simpleUpdateWrapper(classifyDict, classifyDb, genome,
classifier.__name__)
detailsDict = pickle.load(
open(
os.path.join(self.tmpDir, genome,
"Details" + classifier.__name__ + genome),
"rb"))
self.simpleBedUpdateWrapper(detailsDict, detailsDb, genome,
classifier.__name__)
def augustusNotOk():
"""
Defines augustus OK as not failing all Augustus classifiers.
"""
classifyFields = detailsFields = [x.__name__ for x in classes_in_module(src.augustusClassifiers)]
classifyOperations = ["OR"] * (len(classifyFields) - 1)
classifyValues = [1] * (len(classifyFields))
return detailsFields, classifyFields, classifyValues, classifyOperations
def run(self):
classifiers = classes_in_module(src.classifiers)
attributes = classes_in_module(src.attributes)
classifyDb = os.path.join(self.outDir, "classify.db")
detailsDb = os.path.join(self.outDir, "details.db")
attributesDb = os.path.join(self.outDir, "attributes.db")
self.initializeDb(classifyDb, classifiers, dataType="INTEGER")
self.initializeDb(detailsDb, classifiers, dataType="TEXT")
for classifier, genome in product(classifiers, self.genomes):
classifyDict = pickle.load(open(os.path.join(self.tmpDir, genome, "Classify" + classifier.__name__ + genome), "rb"))
self.simpleUpdateWrapper(classifyDict, classifyDb, genome, classifier.__name__)
detailsDict = pickle.load(open(os.path.join(self.tmpDir, genome, "Details" + classifier.__name__ + genome), "rb"))
self.simpleBedUpdateWrapper(detailsDict, detailsDb, genome, classifier.__name__)
self.initializeDb(attributesDb, attributes)
for attribute, genome in product(attributes, self.genomes):
attributeDict = pickle.load(open(os.path.join(self.tmpDir, genome, "Attribute" + attribute.__name__ + genome), "rb"))
self.simpleUpdateWrapper(attributeDict, attributesDb, genome, attribute.__name__)
def run_tm_classifiers(args, target, tmp_dir):
tm_classifiers = classes_in_module(src.alignment_classifiers)
for classifier in tm_classifiers:
target.addChildTarget(
classifier(args.refFasta, args.annotationGp, args.refGenome,
tmp_dir, args.genome, args.psl, args.refPsl, args.fasta,
args.targetGp))
attributes = classes_in_module(src.attributes)
for attribute in attributes:
target.addChildTarget(
attribute(args.refFasta, args.annotationGp, args.refGenome,
tmp_dir, args.genome, args.psl, args.refPsl, args.fasta,
args.targetGp, args.gencodeAttributes))
# in transMap mode we run the alignment-free classifiers on the target genome
ref_classifiers = classes_in_module(src.classifiers)
for classifier in ref_classifiers:
target.addChildTarget(
classifier(args.fasta, args.targetGp, args.genome, tmp_dir))
def augustusNotOk():
"""
Defines augustus OK as not failing all Augustus classifiers.
"""
classifyFields = detailsFields = [
x.__name__ for x in classes_in_module(src.augustusClassifiers)
]
classifyOperations = ["OR"] * (len(classifyFields) - 1)
classifyValues = [1] * (len(classifyFields))
return detailsFields, classifyFields, classifyValues, classifyOperations
def run(self):
augustusClassifiers = classes_in_module(src.augustusClassifiers)
classifyDb = os.path.join(self.outDir, "augustusClassify.db")
if os.path.exists(classifyDb):
os.remove(classifyDb)
detailsDb = os.path.join(self.outDir, "augustusDetails.db")
if os.path.exists(detailsDb):
os.remove(detailsDb)
self.initializeDb(classifyDb, augustusClassifiers, dataType="INTEGER")
self.initializeDb(detailsDb, augustusClassifiers, dataType="TEXT")
for classifier, genome in product(augustusClassifiers, self.genomes):
classifyDict = pickle.load(open(os.path.join(self.tmpDir, genome, "Classify" + classifier.__name__ + genome), "rb"))
self.simpleUpdateWrapper(classifyDict, classifyDb, genome, classifier.__name__)
detailsDict = pickle.load(open(os.path.join(self.tmpDir, genome, "Details" + classifier.__name__ + genome), "rb"))
self.simpleBedUpdateWrapper(detailsDict, detailsDb, genome, classifier.__name__)
def allProblems():
detailsFields = classifyFields = [x.__name__ for x in classes_in_module(src.classifiers)]
classifyOperations = ["OR"] * (len(classifyFields) - 1)
classifyValues = [1] * len(classifyFields)
return detailsFields, classifyFields, classifyValues, classifyOperations
def run_ref_classifiers(args, target, tmp_dir):
ref_classifiers = classes_in_module(src.classifiers)
for classifier in ref_classifiers:
target.addChildTarget(
classifier(args.refFasta, args.annotationGp, args.refGenome,
tmp_dir))
