def testParserBigFile(self):
# ps = PositiveSelectionParser
# exp = Consume(ps.selection_header).suppress()+OneOrMore(ps.selection_line)+empty
# exp.setDebug(True)
# with open(self.file+"1") as handle:
# r = exp.parseFile(handle)
# print "result",r
parser = PositiveSelectionParser()
with open(self.file) as handle:
models = list(parser.parse(handle))
assert len(models)==3
index = {}
for model in models:
index[model.model] = sorted([(ps.column, ps.probability) for ps in model.positive_selection])
assert len(index[2]) == 2
assert index[2] == [(188,float("0.573")),
(193,float("0.560"))]
assert len(index[8]) == 3
assert index[8][-1] == (582,float("0.932"))
print index[3]
assert len(index[3]) == 4
assert index[3] == [(2,float("0.925")),
(3,float("0.998")),
(4,float("0.917")),
(582,float("1.0"))]
def _codeml(self):
if not self.codemlFile:
return
assert self.family.id != None
assert self.tree.id != None
# We need to convert the columns to the original alignment indices
mapper = CulledColumnMapper(self.alignment,
self.alignment.culled_columns)
parser = PositiveSelectionParser()
models = list(parser.parse(self.codemlFile))
runtime().debug("Found", len(models), "models")
for i, model in enumerate(models):
model.tree_key = self.tree.id
self.session.add(model)
self.session.flush()
ps = list(model.ps)
runtime().debug("Found", len(ps), "sites in model", model.model)
for j, site in enumerate(ps):
site.codeml_key = model.id
# Indices in CodeML start at 1, convert to 0 and then map
orig = site.column
site.column = mapper[site.column - 1]
runtime().debug("column", orig, "mapped to", site.column,
site.probability)
try:
self.session.add(site)
except:
runtime().debug(i, ":", j, " failure on column", orig,
"mapped to", site.column, site.probability)
raise
runtime().debug("Finished with model")
self.session.flush()
# with open(self.codemlFile) as handle:
# text = handle.read()
# from hpf.hddb.db import CodeML
# self.codeml = CodeML(tree_key=self.tree.id,
# filename=self.codemlFile,
# text=text)
# self.session.add(self.codeml)
# self.session.flush()
# parser = LRTParser(self.alignment, self.alignment.culled_columns,self.codeml)
# with open(self.codemlFile) as handle:
# for selection in parser.parse(handle):
# selection.codeml_key = self.codeml.id
# self.session.merge(selection)
runtime().debug("finished import codeml")
def _codeml(self):
if not self.codemlFile:
return
assert self.family.id != None
assert self.tree.id != None
# We need to convert the columns to the original alignment indices
mapper = CulledColumnMapper(self.alignment, self.alignment.culled_columns)
parser = PositiveSelectionParser()
models = list(parser.parse(self.codemlFile))
runtime().debug("Found", len(models), "models")
for i, model in enumerate(models):
model.tree_key = self.tree.id
self.session.add(model)
self.session.flush()
ps = list(model.ps)
runtime().debug("Found", len(ps), "sites in model", model.model)
for j, site in enumerate(ps):
site.codeml_key = model.id
# Indices in CodeML start at 1, convert to 0 and then map
orig = site.column
site.column = mapper[site.column - 1]
runtime().debug("column", orig, "mapped to", site.column, site.probability)
try:
self.session.add(site)
except:
runtime().debug(i, ":", j, " failure on column", orig, "mapped to", site.column, site.probability)
raise
runtime().debug("Finished with model")
self.session.flush()
# with open(self.codemlFile) as handle:
# text = handle.read()
# from hpf.hddb.db import CodeML
# self.codeml = CodeML(tree_key=self.tree.id,
# filename=self.codemlFile,
# text=text)
# self.session.add(self.codeml)
# self.session.flush()
# parser = LRTParser(self.alignment, self.alignment.culled_columns,self.codeml)
# with open(self.codemlFile) as handle:
# for selection in parser.parse(handle):
# selection.codeml_key = self.codeml.id
# self.session.merge(selection)
runtime().debug("finished import codeml")
def main(fam, t, *files):
global family, tree, session
session = Session()
family = int(fam)
tree = int(t)
for file in files:
#runtime().set_debug(1)
runtime().debug("Using file", file)
with open(file) as handle:
ps = PositiveSelectionParser().parse(handle)
count = consume(imap(merge, ps))
runtime().debug("Found", count, "sites")
session.commit()
session.close()
def testParserSmallFile(self):
parser = PositiveSelectionParser()
with open(self.file+"1") as handle:
models = list(parser.parse(handle))
assert len(models)==1
assert len(models[0].positive_selection)==4