def script(path, out):
omim_dict = get_hp_ic.parse_anno(
open('/dupa-filer/talf/diff-diagnosis/phenotype_annotation.tab'))
diseases = list(omim_dict.keys())
ic_dict = get_hp_ic.load_ic(open('ic/ic_parent.txt'))
contents = os.listdir(path)
patients = [f for f in contents if f.endswith('_hpo.txt')]
t1 = time.time()
for i, p in enumerate(patients):
print(i, time.time() - t1)
try:
phenotypes = open(os.path.join(path, p)).readlines()[0].split(',')
except IndexError:
continue
omim_scores = []
for o, phe in omim_dict.items():
dist = get_hp_ic.calc_dis_sim_no_norm(
omim.Disease(None, None, None, phe),
omim.Disease(None, None, None, phenotypes), ic_dict)
omim_scores.append((o, dist))
omim_scores.sort(key=lambda x: x[1], reverse=True)
with open(os.path.join(out, p + '.results'), 'w') as f:
for line in omim_scores[:20]:
f.write('\t'.join(['1.000',
str(line[1]), 'OMIM:' + line[0]]) + '\n')
def script(path,ic, out):
omim_dict = get_hp_ic.parse_anno(open('/dupa-filer/talf/diff-diagnosis/phenotype_annotation.tab'))
#diseases = list(omim_dict.keys())
ic_dict = get_hp_ic.load_ic(open(ic))
contents = os.listdir(path)
patients = [f for f in contents if f.endswith('.txt')]
fail_counter = 0
failed_keys = set()
for i, p in enumerate(patients):
if i % 10 == 0: print(i)
try:
phenotypes = open(os.path.join(path, p)).readlines()[0].split(',')
except IndexError:
continue
omim_scores = []
for o, phe in omim_dict.items():
try:
sim = get_hp_ic.calc_simgic(omim.Disease(None, None, None, phe), omim.Disease(None, None, None, phenotypes), ic_dict)
except KeyError as e:
fail_counter += 1
failed_keys.add(e.args[0])
omim_scores.append((o, sim))
omim_scores.sort(key=lambda x: x[1],reverse=True)
with open(os.path.join(out, p + '.results'), 'w') as f:
for line in omim_scores[:20]:
f.write('\t'.join(['1.000', str(line[1]), line[0]]) + '\n')
print("Number of occurences of missing phenotypes: " + str(fail_counter) + '\n')
print("Number of unique phenos failing: " + str(len(failed_keys)) + '\n')
print(failed_keys)
def script(path,ic, out, pvals, normalized):
omim_dict = get_hp_ic.parse_anno(open('/dupa-filer/talf/diff-diagnosis/phenotype_annotation.tab'))
#diseases = list(omim_dict.keys())
if normalized == 'half':
normalized = 'half'
elif normalized == 'True':
normalized = True
elif normalized == 'False':
normalized = False
else:
print('Bad option {}'.format(normalized))
sys.exit()
#ic_dict = collections.defaultdict(lambda: 1)
ic_dict = get_hp_ic.load_ic(open(ic))
contents = os.listdir(path)
patients = [f for f in contents if f.endswith('.txt')]
(pval_table, pval_dict) = load_pvals(pvals)
fail_counter = 0
failed_keys = set()
t1 = time.time()
for i, p in enumerate(patients):
if i % 10 == 0: print(i, time.time())
try:
phenotypes = open(os.path.join(path, p)).readlines()[0].split(',')
except IndexError:
continue
omim_scores = []
for o, phe in omim_dict.items():
if not o in pval_dict: continue
if len(phenotypes) == 0: (pval, sim, final_score) = (1.0, 0, 0)
else:
try:
pval_list = pval_table[pval_dict[o]][min(len(phenotypes) - 1, 9)]
#norm True will give intersection/union, norm False just gives straight count
sim = get_hp_ic.calc_simgic(omim.Disease(None, None, None, phenotypes), omim.Disease(None, None, None, phe), ic_dict, ancestors=True, norm=normalized)
pval = get_hp_ic.get_p_value(sim, pval_list) * 0.5
final_score = (1 - pval) * 1000000000000 + sim
except KeyError as e:
fail_counter += 1
failed_keys.add(e.args[0])
omim_scores.append((o, final_score, pval, sim))
omim_scores.sort(key=lambda x: x[1],reverse=True)
with open(os.path.join(out, p + '.results'), 'w') as f:
for line in omim_scores[:20]:
f.write('\t'.join([str(line[2]), str(line[3]), line[0]]) + '\n')
print("Number of occurences of missing phenotypes: " + str(fail_counter) + '\n')
print("Number of unique phenos failing: " + str(len(failed_keys)) + '\n')
print(failed_keys)
def script(path,ic, out, pvals):
omim_dict = get_hp_ic.parse_anno(open('/dupa-filer/talf/diff-diagnosis/phenotype_annotation.tab'))
#diseases = list(omim_dict.keys())
ic_dict = get_hp_ic.load_ic(open(ic))
contents = os.listdir(path)
patients = [f for f in contents if f.endswith('.txt')]
(pval_table, pval_dict) = load_pvals(pvals)
fail_counter = 0
failed_keys = set()
t1 = time.time()
for i, p in enumerate(patients):
if i % 10 == 0: print(i, time.time() - t1)
try:
phenotypes = open(os.path.join(path, p)).readlines()[0].split(',')
except IndexError:
continue
omim_scores = []
for o, phe in omim_dict.items():
if not o in pval_dict: continue
if len(phenotypes) == 0:
(pval, sim, final_score) = (1.0, 0, 0)
else:
pval_list = pval_table[pval_dict[o]][min(len(phenotypes) - 1, 9)]
try:
sim = get_hp_ic.calc_simgic(omim.Disease(None, None, None, phe), omim.Disease(None, None, None, phenotypes), ic_dict)
pval = get_hp_ic.get_p_value(sim, pval_list) * 0.1
final_score = (1 - pval) * 1000000000000 + sim
except KeyError as e:
fail_counter += 1
failed_keys.add(e.args[0])
omim_scores.append((o, final_score, pval, sim))
omim_scores.sort(key=lambda x: x[1], reverse=True)
with open(os.path.join(out, p + '.results'), 'w') as f:
for line in omim_scores[:20]:
f.write('\t'.join([str(line[2]), str(line[3]), line[0]]) + '\n')
print("Number of occurences of missing phenotypes: " + str(fail_counter) + '\n')
print("Number of unique phenos failing: " + str(len(failed_keys)) + '\n')
print(failed_keys)
def script(path, out):
omim_dict = get_hp_ic.parse_anno(open('/dupa-filer/talf/diff-diagnosis/phenotype_annotation.tab'))
diseases = list(omim_dict.keys())
ic_dict = get_hp_ic.load_ic(open('ic/ic_parent.txt'))
contents = os.listdir(path)
patients = [f for f in contents if f.endswith('_hpo.txt')]
t1 = time.time()
for i, p in enumerate(patients):
print(i, time.time() - t1)
try:
phenotypes = open(os.path.join(path, p)).readlines()[0].split(',')
except IndexError:
continue
omim_scores = []
for o, phe in omim_dict.items():
dist = get_hp_ic.calc_dis_sim_no_norm(omim.Disease(None, None, None, phe), omim.Disease(None, None, None, phenotypes), ic_dict)
omim_scores.append((o, dist))
omim_scores.sort(key=lambda x: x[1],reverse=True)
with open(os.path.join(out, p + '.results'), 'w') as f:
for line in omim_scores[:20]:
f.write('\t'.join(['1.000', str(line[1]), 'OMIM:' + line[0]]) + '\n')
import sys
import get_hp_ic
import random
import omim
import time
import hpo_lib
if __name__ == '__main__':
dis_code = sys.argv[1]
outdir = sys.argv[2]
ic_file = sys.argv[3]
dis_dict = get_hp_ic.parse_anno(
open('/dupa-filer/talf/diff-diagnosis/phenotype_annotation.tab',
encoding='utf-8'))
pheno_dist = []
for d in dis_dict:
valid = filter(lambda hp: hpo_lib.search_code(hp[3:]) != None,
dis_dict[d])
pheno_dist += valid
dis_phenos = dis_dict[dis_code]
dis = omim.Disease(None, None, None, dis_phenos)
score_fn = get_hp_ic.calc_simgic
random_fn = random.choice
omim_cons = omim.Disease
ic_dict = get_hp_ic.load_ic(open(ic_file))
t1 = time.time()
for query_size in range(1, 11):
scores = []
for trial in range(100000):
if trial % 1000 == 0: print(trial, time.time() - t1)
query = [random_fn(pheno_dist) for i in range(query_size)]
import sys
import get_hp_ic
import random
import omim
import time
if __name__ == '__main__':
dis_code = sys.argv[1]
outdir = sys.argv[2]
ic_file = sys.argv[3]
dis_dict = get_hp_ic.parse_anno(open('/dupa-filer/talf/diff-diagnosis/phenotype_annotation.tab', encoding='utf-8'))
pheno_dist = []
for d in dis_dict:
pheno_dist += dis_dict[d]
dis_phenos = dis_dict[dis_code]
dis = omim.Disease(None, None, None, dis_phenos)
score_fn = get_hp_ic.calc_simgic
random_fn = random.choice
omim_cons = omim.Disease
ic_dict = get_hp_ic.load_ic(open(ic_file))
t1 = time.time()
for query_size in range(1, 11):
scores = []
for trial in range(100000):
if trial % 1000 == 0: print(trial, time.time() - t1)
query = [random_fn(pheno_dist) for i in range(query_size)]
score = score_fn(omim_cons(None, None, None, query), dis, ic_dict)
scores.append(score)
scores.sort()
fout = open('{}/{}.txt'.format(outdir, str(query_size)), 'w')
for score in scores: