class GoFastSimilarity(GoSimilarity):
# the biological process root node
BP_root = 8150
def __init__(self, obo_file, sql_conn, verbose=True):
# load gene ontology with fastSemSim
if verbose:
print("loading gene ontology")
# import only the BP namespace:
self.go_dag = GODag(obo_file, only_namespace="biological_process")
# load annotation class with fastSemSim
if verbose:
print("loading associations")
self.assoc = self.load_go_associations(sql_conn)
if verbose:
print("initializing GO Dag probabilities and IC")
self.go_dag.term_probability(self.assoc)
self.go_dag.term_IC()
def _simRel_score(self, term1, term2):
# get the MICA (common ancestor with maximal IC)
max_ic_anc = self.go_dag.get_max_IC_anc(term1, term2)
# get IC and p() of that term
lca_IC = self.go_dag.IC[max_ic_anc]
lca_p = self.go_dag.p[max_ic_anc]
# calc the denominator for the score
denom = self.go_dag.IC[term1] + self.go_dag.IC[term2]
if denom != 0:
score = 2 * lca_IC / denom * (1 - lca_p)
else:
score = 0.0
if score > 1.0 or score < 0.0:
raise Exception("SimRel score is invalid")
return score
def term_pairwise_score(self, term1, term2):
return self._simRel_score(term1, term2)
def gene_pairwise_score(self, gene1, gene2):
# get associated GO-Terms:
terms1 = self.assoc[gene1]
terms2 = self.assoc[gene2]
# get scores:
scores = self.terms_sets_scores(terms1, terms2)
# get maximum score
if len(scores) == 0:
# set to zero in case there are no scores
score = 0.0
else:
score = max(scores)
if score > 1.0 or score < 0.0:
raise Exception
return score
def __init__(self, obo_file, sql_conn, verbose=True):
# load gene ontology with fastSemSim
if verbose:
print("loading gene ontology")
# import only the BP namespace:
self.go_dag = GODag(obo_file, only_namespace="biological_process")
# load annotation class with fastSemSim
if verbose:
print("loading associations")
self.assoc = self.load_go_associations(sql_conn)
if verbose:
print("initializing GO Dag probabilities and IC")
self.go_dag.term_probability(self.assoc)
self.go_dag.term_IC()