def get_greedy_communities():
greedy_pickle_path = 'CachedFiles/greedy_communities.pickle'
if path.exists(greedy_pickle_path):
print('Loading Greedy Communities')
return pickle.load(open(greedy_pickle_path, 'rb'))
else:
print('Generating Greedy Communities')
G, Gn = load_graphs()
c = list(nx.algorithms.community.modularity_max.greedy_modularity_communities(G))
# convert from 2D frozenset to 2D list
return [[y for y in x] for x in c]
def get_reduced_gtex(reactome):
filepath = 'CachedFiles/gtex-pandas-mygene2-genes-only.pickle'
global GTEX_GLOBAL
if path.exists(filepath) and False:
print('Loading Reduced GTEX')
GTEX_GLOBAL = None
return pickle.load(open(filepath, 'rb'))
else:
print('Generating Reduced GTEX')
exp = get_expression_for_all_genes(reactome)
# load mygene2
G, Gn = load_graphs()
genes = [x for x in G.nodes if x[0:3] != 'HP:']
exp.index = exp['Description']
GTEX_GLOBAL = None
exp = exp.loc[genes]
exp = remove_filler_rows(exp)
pickle.dump(exp, open(filepath, 'wb'), protocol=4)
return exp
hpos_per_gene.append(len(set(hpos)))
print('\n')
print(name + ' Found in StringDB: ' + str(given_graph_found_count))
print(name + ' Not Found in StringDB: ' + str(given_graph_not_found_count))
# the vast majority of StringDB neighbors are not found in MyGene2
print('Neighbors Found: ' + str(found_neighbor_count))
print('Neighbors Not Found: ' + str(not_found_neighbor_count))
return hpos_per_gene, [given_graph_found_count,
given_graph_not_found_count,
found_neighbor_count,
not_found_neighbor_count]
if __name__ == "__main__":
print('Node Expansion Analysis')
G, Gn = load_graphs()
Gj = load_jenkins_gene_to_pheno()
# get the number of HPOs associated with each MyGene2 gene
mg2_hpos_counts = get_hpos_per_gene(G)
print('MyGene2 average number of HPOs per gene: ' + str(np.mean(np.array(mg2_hpos_counts))))
# get the number of HPOs associated with each Jenkins gene
jenkins_hpos_counts = get_hpos_per_gene(Gj)
print('Jenkins average number of HPOs per gene: ' + str(np.mean(np.array(jenkins_hpos_counts))))
# get the number of HPOs associated with each MyGene2 gene after expansion with StringDB data
mg2_expanded_hpo_counts, mg2_stats = get_hpos_per_expanded_gene(G, 'MyGene2', False)
print('MyGene2 Expanded average number of HPOs per gene: ' + str(np.mean(np.array(mg2_expanded_hpo_counts))))
# get the number of HPOs associated with each expanded Jenkins gene
jenkins_expanded_hpos_counts, jenkins_stats = get_hpos_per_expanded_gene(Gj, 'Jenkins', False)
print('Jenkins Expanded average number of HPOs per gene: ' + str(np.mean(np.array(jenkins_expanded_hpos_counts))))
# produce data