import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "single shortest path", args.filepath,
args.nthreads)
session = helper.init_sas(args.filepath)
tot_elapsed = 0.0
build_elapsed = 0.0
algo_elapsed = 0.0
for v in helper.generate_sources(args.nsources, session):
st = time.time()
session.network.shortestPath(nthreads=args.nthreads,
logLevel="aggressive",
links={"name": "links"},
outWeights={
"name": "out",
"replace": True
},
source=v)
tot_elapsed += time.time() - st
b, a = helper.parse_sas_log()
build_elapsed += b
algo_elapsed += a
build_elapsed = build_elapsed / args.nsources
algo_elapsed = algo_elapsed / args.nsources
import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "spanning forest", args.filepath,
args.nthreads)
session = helper.init_sas(args.filepath)
st = time.time()
session.optnetwork.minSpanTree(nThreads=args.nthreads,
logLevel="aggressive",
links={"name": "links"},
out={"name": "outLinks"})
tot_elapsed = time.time() - st
helper.terminate_sas(session, tot_elapsed)
})))
else:
not_finished_number = len(
list(
qm_collection.find({
'$or': [
ssm_query, ts_refine_query, ts_query, irc_query,
irc_opt_query, irc_equal_query,
insert_reaction_query, ard_query
]
})))
if ard_had_add_number == ard_should_add_number and not_finished_number == 0:
print('Network converged')
target = list(status_collection.find({}))
update_field = {'status': 'Network converged'}
status_collection.update_one(target[0], {"$set": update_field}, True)
print_header()
check_ard_jobs()
qm_collection = db['qm_calculate_center']
max_gen = qm_collection.find_one(sort=[("generations", -1)])
max_gen = max_gen['generations']
for i in range(max_gen):
print_information(i + 1)
update_network_status()
import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "local clustering coefficient",
args.filepath, args.nthreads)
session = helper.init_sas(args.filepath)
st = time.time()
session.network.centrality(clusteringCoef=True,
nThreads=args.nthreads,
logLevel="aggressive",
links={"name": "links"},
outNodes={
"name": "out",
"replace": True
}),
tot_elapsed = time.time() - st
helper.terminate_sas(session, tot_elapsed)
def main():
"""Estimates the proportion of causals from GWAS summary statistics
Given summary statistics (effect sizes) from GWAS and an estimate of
genome-wide heritability (H_gw), MCMC is used to estimate the proportion
of causal variants. An efficient implementation of the algorithm allows
for runtime of O(M*K), where M=#SNPs and K=#causalSNPs
Args:
H_gw: estimated genome-wide heritability
M_gw: number of SNPs genome-wide
N: sample size of study
ld_half_file: pre-computed 1/2 power of LD matrix; txt or .npy file
gwas_file: transformed GWAS effect sizes created by left-multplying
effects by V^-1/2; effects must be under header 'BETA_STD_I';
txt file
outdir: path to directory for results files
seed: seed for random initialization
id: string name for experiment
its: number of MCMC iterations
dp: boolean flag; use efficient update if flag present
profile: boolean flag; profile code if flag present
Returns:
An output file containing posterior mean and posterior variance of
the estimate of the proportion of causal variants, as well as the
average likelihood and variance of the likelihood.
example output file:
Estimate p: 0.0020
SD p: 0.0004606
Avg log like: 1452.92
Var log like: 2.12
"""
logging.basicConfig(format='%(asctime)s - %(message)s',
datefmt='%d-%b-%y %H:%M:%S',
level=logging.INFO)
parser = OptionParser()
parser.add_option("--H_gw", type=float)
parser.add_option("--M_gw", type=int)
parser.add_option("--N", type=float)
parser.add_option("--ld_half_file")
parser.add_option("--gwas_file")
parser.add_option("--outdir")
parser.add_option("--seed", type=int, default="2019")
parser.add_option("--id")
parser.add_option("--its", type=int)
parser.add_option("--dp", action="store_true", default=False)
parser.add_option("--profile", action="store_true")
(options, args) = parser.parse_args()
# set seed
seed = options.seed
np.random.seed(seed)
H_gw = options.H_gw
if H_gw < 0:
print "ERROR: estimate of h2 is negative...exiting"
exit(1)
M_gw = options.M_gw
N = options.N
id = options.id
its = options.its
ld_half_file = options.ld_half_file
gwas_file = options.gwas_file
outdir = options.outdir
dp_flag = options.dp
profile_flag = options.profile
# open output filehandler
outfile = os.path.join(outdir, id + '.' + str(seed) + ".unity_v3.log")
f = open(outfile, 'w')
print_header(id, H_gw, N, its, seed, gwas_file, ld_half_file, outdir, f)
try: # txt file
V_half = np.loadtxt(ld_half_file)
except: # npy file
V_half = np.load(ld_half_file)
gwas = pd.read_table(gwas_file, sep=' ')
z = np.asarray(gwas['BETA_STD_I'])
p_est, p_var, avg_log_like, var_log_like, accept_percent = gibbs_sampler_Hgw(
z, H_gw, M_gw, N, V_half, its, f, dp_flag, profile_flag)
accept_percent = accept_percent * 100
print_func("Accept percent: %.4f" % accept_percent, f)
print_func("Estimate p: %.4g" % p_est, f)
print_func("SD p: %.4g" % math.sqrt(p_var), f)
print_func("Avg log like: %.6g" % avg_log_like, f)
print_func("Var log like: %.4g" % math.sqrt(var_log_like), f)
f.close()
import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "closeness centrality", args.filepath,
args.nthreads)
session = helper.init_sas(args.filepath)
if args.weighted:
session.alterTable(name="links",
columns=[{
"name": "weight",
"rename": "weight2"
}])
st = time.time()
session.network.centrality(
nthreads=args.nthreads,
logLevel="aggressive",
linksVar={"auxweight": "weight2"} if args.weighted else {},
links={"name": "links"},
close="weight" if args.weighted else "unweight",
closeNoPath="zero",
outNodes={
"name": "outNodes",
"replace": True
})
tot_elapsed = time.time() - st
helper.terminate_sas(session, tot_elapsed)
import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "kcore", args.filepath, args.nthreads)
session = helper.init_sas(args.filepath)
st = time.time()
session.network.core(
nThreads=args.nthreads,
logLevel="aggressive",
links={"name": "links"},
outNodes={"name": "out", "replace": True})
tot_elapsed = time.time() - st
helper.terminate_sas(session, tot_elapsed)
import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "strongly connected components",
args.filepath, args.nthreads)
session = helper.init_sas(args.filepath)
st = time.time()
session.network.connectedComponents(direction="directed",
nThreads=args.nthreads,
logLevel="aggressive",
links={"name": "links"},
outNodes={
"name": "out",
"replace": True
})
tot_elapsed = time.time() - st
helper.terminate_sas(session, tot_elapsed)
import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "pagerank", args.filepath, args.nthreads)
session = helper.init_sas(args.filepath)
st = time.time()
session.network.centrality(selfLinks=True,
nThreads=args.nthreads,
logLevel="aggressive",
pagerank="unweight",
pagerankAlpha=0.85,
pagerankTolerance=1e-9,
links={"name": "links"},
outNodes={
"name": "out",
"replace": True
})
tot_elapsed = time.time() - st
helper.terminate_sas(session, tot_elapsed)
import time
import helper
args = helper.take_input()
helper.print_header("sas-network", "jaccard similarity", args.filepath,
args.nthreads)
session = helper.init_sas(args.filepath)
st = time.time()
session.network.nodesimilarity(jaccard=True,
nThreads=args.nthreads,
logLevel="aggressive",
links={"name": "links"},
outsimilarity={
"name": "outSim",
"replace": True
})
tot_elapsed = time.time() - st
helper.terminate_sas(session, tot_elapsed)