def collect_collections(organism, organism_file, given_colls, coll_file, all_colls):
"""
Collects all available collections as a list
and returns together with number of unique genes in all
"""
colls = []
genes = []
if organism_file is None:
colls_dir = os.path.join(
get_resources_dir(),
'collections',
organism.split('_')[0]
)
for coll in all_colls:
coll_path = os.path.join(colls_dir, coll + '.json')
if os.path.exists(coll_path):
with open(coll_path, 'r') as f: _coll = json.load(f)
genes.extend([gene_set['genes'] for gene_set in _coll['geneSets']])
if coll in given_colls:
colls.append(_coll)
if coll_file is not None:
_coll = _read_gmt_file(coll_file)
genes.extend([gene_set['genes'] for gene_set in _coll['geneSets']])
colls.append(_coll)
# count the number of unique genes in all collections
colls_size = len(set(chain.from_iterable(genes)))
return colls, colls_size
def generate(organism, organism_file):
"""
Generates an interval tree of mapping from downloaded file
"""
# parse either JSON file or given Tab-separated file as mapping
mapping = []
if organism_file is None:
# Read mapping to a list
mapping_path = path.join(get_resources_dir(), 'mappings', organism + '.json')
if path.exists(mapping_path):
# Collect chromosomes in a dictionary as an interval tree
with open(mapping_path, 'r') as f: mapping = json.load(f)
else:
with open(organism_file, 'r') as rows:
for row in rows:
cols = row.strip().split('\t')
if len(cols) == 4:
mapping.append({
'chrName': cols[0],
'start': cols[1],
'end': cols[2],
'symbol': cols[3],
})
# generate mapping interval tree for fast lookup
mappingTree = {}
for each in mapping:
# have we added that chromosome name before?
if not mappingTree.has_key(each['chrName']):
mappingTree[each['chrName']] = IntervalTree()
mappingTree[each['chrName']].addi(
int(each['start']), int(each['end']), each['symbol'])
return mappingTree
