def main(args):
print("Loading SPM data...")
spm = flower.load_spm(args.spm_file)
nodelist = spm['nodelist']
srcmodules = utils.load_nodeset_file(args.srcmodules_file, sep=args.sep)
destmodules = utils.load_nodeset_file(args.destmodules_file, sep=args.sep)
if not isinstance(nodelist[0], str):
srcmodules = list(map(lambda x: list(map(int, x)), srcmodules))
destmodules = list(map(lambda x: list(map(int, x)), destmodules))
print("Evaluating flows...")
flows_batch = flower.evaluate_flows_batch_nx(srcmodules,
destmodules,
spm,
n_threads=args.n_threads)
print("Completed.")
print("Writing to file...")
flows_batch.to_csv(args.output_file, sep='\t')
print("Completed.")
def main(args):
print("Loading SPM data...")
spm_data = flower.load_spm(args.spm_file)
srcnodes = utils.load_nodeset_file(args.srcnodes_file,
sep=args.sep,
as_column=args.as_col)
destnodes = utils.load_nodeset_file(args.destnodes_file,
sep=args.sep,
as_column=args.as_col)
print(len(srcnodes), len(destnodes))
if not isinstance(spm_data['nodelist'][0], str):
srcnodes = list(map(int, srcnodes))
destnodes = list(map(int, destnodes))
print("Evaluating flows...")
flows = flower.eval_flow_centrality_nx(srcnodes, destnodes, spm=spm_data)
print("Completed.")
print(flows.iloc[:10])
print("Writing to file...")
flows.to_csv(args.output_file, sep='\t')
print("Completed.")
def main(args):
if args.rdm_fc_path is None and (args.rdm_src_path is None
or args.rdm_dest_path is None):
raise ValueError(
'Either rdm_fc_path or (rdm_src_path, rdm_dest_path) have to be specified!'
)
print("Loading SPM data...")
spm = flower.load_spm(args.spm_file)
nodelist = spm['nodelist']
srcnodes = utils.load_nodeset_file(args.srcnodes_file,
sep=args.sep,
as_column=args.as_col)
destnodes = utils.load_nodeset_file(args.destnodes_file,
sep=args.sep,
as_column=args.as_col)
if not isinstance(nodelist[0], str):
srcnodes = list(map(int, srcnodes))
destnodes = list(map(int, destnodes))
print("Evaluating flows...")
if args.rdm_fc_path is None:
rdm_srcmodules = utils.load_nodeset_file(args.rdm_src_path,
sep=args.sep)
rdm_destmodules = utils.load_nodeset_file(args.rdm_dest_path,
sep=args.sep)
if not isinstance(nodelist[0], str):
rdm_srcmodules = map(lambda x: map(int, x), rdm_srcmodules)
rdm_destmodules = map(lambda x: map(int, x), rdm_destmodules)
fcs = flower.evaluate_fcs_nx(srcnodes,
destnodes,
rdm_srcmodules,
rdm_destmodules,
spm,
fill_na=args.fill_na,
n_threads=args.n_threads)
else:
rdm_fc = pd.read_csv(args.rdm_fc_path, sep='\t', index_col=0)
fc = flower.eval_flow_centrality_nx(srcnodes, destnodes, spm)
fcs = flower.evaluate_flows_statistics_from_batch_nx(rdm_fc, fc)
fcs = flower.enrich_flows_table_nx(srcnodes,
destnodes,
fcs,
spm,
fill_na=args.fill_na)
print("Completed.")
print("Writing to file...")
fcs.to_csv(args.output_file, sep=args.sep)
print("Completed.")
def main(args):
print ("Loading SPM data...")
spm = flower.load_spm(args.spm_file)
nodelist = spm['nodelist']
srcmodules = pd.read_csv(args.srcmodules_file, sep=args.sep).values.tolist()
destmodules = pd.read_csv(args.destmodules_file, sep=args.sep).values.tolist()
if type(nodelist[0]) is int:
srcmodules = map(lambda x: map(int,x), srcmodules)
destmodules = map(lambda x: map(int,x), destmodules)
print ("Evaluating flows...")
flows_batch = flower.evaluate_flows_statistics_nx(srcmodules, destmodules, spm, true_flows=, n_threads=args.n_threads)
print ("Completed.")
print ("Writing to file...")
flows_batch.to_csv(args.output_file, sep='\t')
print ("Completed.")
flows['FCScore'] = (flows.Flow_value.values - flows.Flow_mean.values) / flows.Flow_std.values flows[np.isnan(flows.FCScore) | np.isinf(flows.FCScore)].FCScore = None print "Creating rich columns..." # NOTE: this operation cannot be parallelized when gene symbols have to be retrieved on mygene server, probably the sock connection to the mygene server doesn't work in parallel flows['Degree'] = flows.apply(lambda row: nx.degree(ppi, row.NodeID),axis=1) flows['in_DestModule'] = flows.apply(lambda row: row.NodeID in destnodes,axis=1) flows['in_SrcModule'] = flows.apply(lambda row: row.NodeID in srcnodes,axis=1) print "Calculating distance statistics..." spm = flower.load_spm(args.spm_file) gm = utils.GIDMapper(spm['nodelist']) destnodes_ids = gm.gid2id(destnodes) srcnodes_ids = gm.gid2id(srcnodes) ids = gm.gid2id(flows.NodeID.values) flows['min_dist_dest'] = flower.eval_min_distance(spm['path_lengths'], destnodes_ids, ids) flows['min_dist_src'] = flower.eval_min_distance(spm['path_lengths'], srcnodes_ids, ids) flows['avg_dist_dest'] = flower.eval_avg_distance(spm['path_lengths'], destnodes_ids, ids) flows['avg_dist_src'] = flower.eval_avg_distance(spm['path_lengths'], srcnodes_ids, ids) print "Sorting..."
default=1,
help='Number of parallel processes')
parser.add_argument('--int_nodeid',
action='store_true',
help='Cast the node IDs to integer')
parser.add_argument('--dry', '--dry_run', action='store_true', help='Dry run')
args = parser.parse_args()
print "Importing network..."
nodetype = int if args.int_nodeid else str
ppi = nx.read_edgelist(args.network_file, nodetype=nodetype)
print "Loading SPM data..."
spm_data = flower.load_spm(args.spm_file)
nodelist = spm_data['nodelist']
with open(args.srcdestmodules_file, 'r') as f:
srcrows, destrows = zip(
*map(lambda x: x.rstrip().split('|'), f.readlines())
) # this line is tricky but it just splits the rows in srcmodule and destmodule
srcmodules = map(lambda x: map(nodetype, x.split(' ')), srcrows)
destmodules = map(lambda x: map(nodetype, x.split(' ')), destrows)
gm = utils.GIDMapper(nodelist=nodelist)
nodes_dest_list = map(gm.gid2id, destmodules)
nodes_src_list = map(gm.gid2id, srcmodules)