def process_ontology(self, obo_file):
obo_dag = GODag(obo_file)
obo_dag.populate_terms()
diseases = self.diseases
ontology_map = {}
self.category_map = defaultdict(list)
for item_id, item in obo_dag.items():
# Considers those diseases whose names are subsets of the disease ontology names
correlated_diseases = [
disease for disease in diseases if disease in item.name
]
if len(correlated_diseases) > 0:
d = {}
for parent in (obo_dag.paths_to_top(item.id)[0]):
d[parent.level] = parent.name
for corr_disease in correlated_diseases:
self.category_map[parent.name] += [corr_disease]
for corr_disease in correlated_diseases:
# Chooses the most specific disease mapping
if corr_disease in ontology_map:
if len(ontology_map[corr_disease]) > len(d):
continue
ontology_map[corr_disease] = d
self.dag = obo_dag
self.disease_class_map = ontology_map
class Data(object):
"""Holds data used in test."""
def __init__(self, fin_obo):
self.repo = os.path.join(os.path.dirname(os.path.abspath(__file__)), "../")
self.fin_obo = os.path.join(self.repo, fin_obo)
self.dag = GODag(self.fin_obo)
self.go2obj = {go:o for go, o in self.dag.items() if not o.is_obsolete}
self.goids_all = self.go2obj.keys()
def get_goids(self, num):
"""Return N randomly chosen GO IDs."""
shuffle(self.goids_all)
return set(self.goids_all[:num])
def get_pathway_mapping(organism=9606, ontology='basic', exclude=None, force=False):
obo = 'goslim_generic.obo' if 'slim' in ontology else 'go-basic.obo'
namespace_filter = get_namespace_filter(exclude)
if force & (os.path.isfile(obo)):
os.remove(obo)
obo_fname = goatools.base.download_go_basic_obo(obo)
obodag = GODag(obo_fname)
return {term_id:term.name for term_id,term in obodag.items() if namespace_filter(term.namespace)}
class Data(object):
"""Holds data used in test."""
def __init__(self, fin_obo):
self.repo = os.path.join(os.path.dirname(os.path.abspath(__file__)),
"../")
self.fin_obo = os.path.join(self.repo, fin_obo)
self.dag = GODag(self.fin_obo)
self.go2obj = {
go: o
for go, o in self.dag.items() if not o.is_obsolete
}
self.goids_all = self.go2obj.keys()
def get_goids(self, num):
"""Return N randomly chosen GO IDs."""
shuffle(self.goids_all)
return set(self.goids_all[:num])
def main():
data = collections.defaultdict(set)
g = GODag()
selection = set()
for name, rec in g.items():
if rec.namespace!="biological_process" or rec.level < 1: continue
selection.add(rec.id)
fp = file("gene_association.tair")
for row in fp:
if row[0]=="!": continue
atoms = row.split("\t")
#['TAIR', 'locus:2185485', 'AT5G14850', '', 'GO:0000030', 'TAIR:Communication:501714663', 'ISS', 'NCBI_gi:1552169|NCBI_gi:7634741', 'F', 'AT5G14850', 'AT5G14850|T9L3.150|T9L3_150', 'protein', 'taxon:3702', '20021003', 'TIGR', '', 'TAIR:locus:2185485\n']
domain, name, go = atoms[0], atoms[10], atoms[4]
name = name.split("|", 1)[0]
if go in selection and domain=="TAIR":
data[name].add(go)
fw = file("microarray.assoc", "w")
print >>fw, "#gene,go_terms"
for key, val in sorted(data.items()):
print >>fw, "%s,%s" % (key, ";".join(sorted(val)))
# syn = TermSynonym(term=dbTerm, synonym=synonym[0][:255])
# syn.save()
# except:
# pass
#
# for synonym in term.alt_ids:
# try:
# syn = TermSynonym(term=dbTerm, synonym=synonym[0][:255])
# syn.save()
# except:
# pass
names = {}
cache = {x.identifier:x for x in Term.objects.filter(ontology=ontology)}
go_dag = GODag("/data/databases/so/so.obo", optional_attrs=['relationship'], load_obsolete=False)
for go, term in tqdm(go_dag.items()):
if term.name in names:
continue
names[term.name] = 1
for child in term.children:
if go in cache and child.id in cache:
r = TermRelationship(
subject_term=cache[go], # parent
predicate_term=is_a,
object_term=cache[child.id], # child
ontology=ontology
' wget http://www.geneontology.org/ontology/subsets/goslim_generic.obo')
obo_fname = download_go_basic_obo()
from goatools.base import download_ncbi_associations
gene2go = download_ncbi_associations()
if goset == 'goslim':
obodag = GODag("goslim_generic.obo")
else:
obodag = GODag("go-basic.obo")
geneid2gos = read_ncbi_gene2go("gene2go", taxids=[9606])
levels = [r.level for go, r in obodag.items()]
[(i, levels.count(i)) for i in range(1, 12)]
bad_go = []
for go, r in obodag.iteritems():
if r.level > cutlvl:
bad_go.append(go)
bad_go = set(bad_go)
len(bad_go)
for go, r in obodag.items():
nps = set()
for p in r._parents:
if p in bad_go:
nps |= set([p2 for p2 in obodag[p]._parents if not p2 in bad_go])
