def read_fsl(filename):
groups = readers.read_grouping('grouping.csv',
short=True,
map_name=True,
strip_colon=False)
bait_to_match = {}
with open(filename) as f:
lines = f.readlines()
skipped_matches = 0
for line in lines[5:]:
tokens = line.strip('\n').split('\t')
score = int(tokens[0])
gene = tokens[9]
bait = tokens[13]
if bait in bait_to_match:
if score > bait_to_match[bait][0]:
bait_to_match[bait] = (score, gene)
else:
bait_to_match[bait] = (score, gene)
total = collections.defaultdict(int)
for s, g in bait_to_match.values():
if g in groups:
total[groups[g]] += 1
else:
total['unknown'] += 1
return total
def main():
if len(sys.argv) != 2:
print("Usage: bait_count.py master_amr_parsed.tsv")
sys.exit(-1)
count_file = sys.argv[1]
name_to_count = read_count(count_file)
groups = readers.read_grouping('grouping.csv', short=True, map_name=True, strip_colon=False)
terms = ontology_common.parse_obo('new_combined.obo')
baits_for_class = collections.defaultdict(int)
for gene, count in name_to_count.items():
group = groups[gene]
for cl in ontology_common.get_class(group, terms):
baits_for_class[cl] += count
print()
total_baits = 0
for k, v in baits_for_class.items():
if 'resistance gene' in terms[k]['name'][0]:
print(terms[k]['name'][0], v)
total_baits += v
print("Total counts for gene class ", total_baits)
print()
total_baits = 0
for k, v in baits_for_class.items():
if 'resistance gene' not in terms[k]['name'][0]:
print(terms[k]['name'][0], v)
total_baits += v
print("Total counts for mechanism ", total_baits)
def main():
if len(sys.argv) != 4:
print("Usage: recluster.py grouping.csv cdhit output.csv")
sys.exit(0)
id_to_name = readers.read_grouping(sys.argv[1], short=True)
new_name_to_id = {}
for id, names in id_to_name.items():
try:
clstr_id_to_name = readers.read_cluster("%s/%s.clstr" %
(sys.argv[2], id))
except IOError:
# work around bug in cdhit where it cannot deal with one sequence
clstr_id_to_name = {0: ['ENA|pgpA/ltpgpA|CAA']}
for clstr_id, names in clstr_id_to_name.items():
new_id = 'ARO:%ss%s' % (id[3:], clstr_id)
for name in names:
new_name_to_id[name] = new_id
with open(sys.argv[1]) as f:
lines = f.readlines()
with open(sys.argv[3], 'w+') as new_grouping:
for line in lines:
line = line.rstrip('\n')
name, id = line.rsplit(',', 1)
short_name = name.strip('"').strip()
short_name = short_name.split(
)[0] if ' ' in short_name else short_name
if len(short_name) > 19:
short_name = short_name[:19]
new_grouping.write('%s,%s\n' % (name, new_name_to_id[short_name]))
def main():
if len(sys.argv) != 4:
print("Usage: recluster.py grouping.csv cdhit output.csv")
sys.exit(0)
id_to_name = readers.read_grouping(sys.argv[1], short=True)
new_name_to_id = {}
for id, names in id_to_name.items():
try:
clstr_id_to_name = readers.read_cluster("%s/%s.clstr" % (sys.argv[2], id))
except IOError:
# work around bug in cdhit where it cannot deal with one sequence
clstr_id_to_name = {0: ['ENA|pgpA/ltpgpA|CAA']}
for clstr_id, names in clstr_id_to_name.items():
new_id = 'ARO:%ss%s' % (id[3:], clstr_id)
for name in names:
new_name_to_id[name] = new_id
with open(sys.argv[1]) as f:
lines = f.readlines()
with open(sys.argv[3], 'w+') as new_grouping:
for line in lines:
line = line.rstrip('\n')
name, id = line.rsplit(',', 1)
short_name = name.strip('"').strip()
short_name = short_name.split()[0] if ' ' in short_name else short_name
if len(short_name) > 19:
short_name = short_name[:19]
new_grouping.write('%s,%s\n' % (name, new_name_to_id[short_name]))
def read_fsl(filename):
groups = readers.read_grouping('grouping.csv', short=True, map_name=True, strip_colon=False)
query_to_target = {}
with open(filename) as f:
lines = f.readlines()
for line in lines[1:]:
tokens = line.strip('\n').split('\t')
group = groups[tokens[1]]
if tokens[0] in query_to_target:
query_to_target[tokens[0]].append((group, tokens[2]))
else:
query_to_target[tokens[0]] = [(group, tokens[2])]
return query_to_target
def read_fsl(filename):
groups = readers.read_grouping('grouping.csv',
short=True,
map_name=True,
strip_colon=False)
query_to_target = {}
with open(filename) as f:
lines = f.readlines()
for line in lines[1:]:
tokens = line.strip('\n').split('\t')
group = groups[tokens[1]]
if tokens[0] in query_to_target:
query_to_target[tokens[0]].append((group, tokens[2]))
else:
query_to_target[tokens[0]] = [(group, tokens[2])]
return query_to_target
def read_fsl(filename):
groups = readers.read_grouping('grouping.csv', short=True, map_name=True, strip_colon=False)
query_to_target = {}
with open(filename) as f:
lines = f.readlines()
skipped_matches = 0
for line in lines[5:]:
tokens = line.strip('\n').split('\t')
if int(tokens[0]) < 120 * .4:
skipped_matches += 1
continue
if tokens[9] in groups:
group = groups[tokens[9]]
else:
group = "Unknown"
if tokens[9] in query_to_target:
query_to_target[tokens[9]].append((group, tokens[13]))
else:
query_to_target[tokens[9]] = [(group, tokens[13])]
print("matches lower that 40%", skipped_matches)
return query_to_target
def read_fsl(filename):
groups = readers.read_grouping('grouping.csv', short=True, map_name=True, strip_colon=False)
bait_to_match = {}
with open(filename) as f:
lines = f.readlines()
skipped_matches = 0
for line in lines[5:]:
tokens = line.strip('\n').split('\t')
score = int(tokens[0])
gene = tokens[9]
bait = tokens[13]
if bait in bait_to_match:
if score > bait_to_match[bait][0]:
bait_to_match[bait] = (score, gene)
else:
bait_to_match[bait] = (score, gene)
total = collections.defaultdict(int)
for s, g in bait_to_match.values():
if g in groups:
total[groups[g]] += 1
else:
total['unknown'] += 1
return total
def read_fsl(filename):
groups = readers.read_grouping('grouping.csv',
short=True,
map_name=True,
strip_colon=False)
query_to_target = {}
with open(filename) as f:
lines = f.readlines()
skipped_matches = 0
for line in lines[5:]:
tokens = line.strip('\n').split('\t')
if int(tokens[0]) < 120 * .4:
skipped_matches += 1
continue
if tokens[9] in groups:
group = groups[tokens[9]]
else:
group = "Unknown"
if tokens[9] in query_to_target:
query_to_target[tokens[9]].append((group, tokens[13]))
else:
query_to_target[tokens[9]] = [(group, tokens[13])]
print("matches lower that 40%", skipped_matches)
return query_to_target
def main():
if len(sys.argv) != 6:
print("Usage: art_test.py grouping.csv scanresults.scan art_out.maf scores.csv genemark.f")
sys.exit(0)
genemark_to_name = read_genemark(sys.argv[5])
id_to_name = readers.read_grouping(sys.argv[1])
thresh = []
fp = []
tp = []
nf = []
contig_to_read = readers.read_maf(sys.argv[3])
for threshold in range(0, 80, 100):
mismatch = []
scan_id_to_name, scan_name_to_id = readers.read_scan_results(threshold, sys.argv[2],protein=True)
readers.change_RF_to_ARO(scan_name_to_id)
scan_name_to_id = substitute_read_name(scan_name_to_id, contig_to_read, genemark_to_name)
found_score = []
true_positive = 0
false_positive = 0
not_found = 0
total = 0
for id in sorted(id_to_name.keys()):
canonical_id = id.split('s')[0]
canonical_id = 'ARO:' + canonical_id.split('O')[1]
for name in id_to_name[id]:
total += 1
ids = search(name, id, scan_name_to_id)
if ids:
if ids[0][0] == canonical_id:
true_positive += 1
found_score.append((ids[0][1], id, name))
else:
if ';' in ids[0][0] and canonical_id in ids[0][0].split(';'):
true_positive += 1
found_score.append((ids[0][1], id, name))
else:
#print("False Positive: %s, %s" % (name, id))
mismatch.append((name, canonical_id, ids[0][0], ids[0][2]))
for i in range(len(ids)):
#print("Attempt %d: %s %f" % (i, ids[i][0], ids[i][1]))
if ids[i][0] == canonical_id:
found_score.append((ids[i][1], id, name))
break
false_positive += 1
else:
#for k,v in contig_to_read.items():
# if name in v:
# #print("Found read %s in %s" % (name, k))
#print("Not Found: %s,%s" % (name, id))
not_found += 1
thresh.append(threshold)
tp.append(true_positive)
fp.append(false_positive)
nf.append(not_found)
print("misclassified %d notfound %d true positive %d of total %d" % (false_positive,
not_found,
true_positive,
total))
print(threshold, true_positive, false_positive, not_found, total)
x = np.array(fp)# false_positive_rate
y = np.array(tp)# true_positive_rate
t = np.array(thresh)
n = np.array(nf)
found_score.sort()
with open(sys.argv[4], 'w') as f:
for score, hmm, name in found_score:
f.write('%s,%d\n' % (hmm, score - 1))
with open('mismatch.csv', 'w') as f:
f.write('Gene,Expected ID,Found ID\n')
for a, b, c, d in mismatch:
f.write('%s,%s,%s\n' % (b, c, a))
test.graph(x, y, t, n)
output = ''.join([gen_random_sequence(prefix_len),
seq,
gen_random_sequence(suffix_len)])
return "> %d:%d?%s?%s\n%s" %(prefix_len, len(seq) + prefix_len, target, name, output)
def unit_test():
pattern = 'AAAAAA'
value = gen_test(pattern, 'adr001', 'test')
lines = value.split('\n')
offset, target, name = lines[0][1:].split('?', 2)
start, end = offset.split(':')
start = int(start)
end = int(end)
if lines[1][start:end] == pattern:
print("Success")
else:
print(lines[1][start:end])
#unit_test()
id_to_name = readers.read_grouping()
name_to_seq = readers.read_fasta()
for id, names in id_to_name.items():
with open('../test/%s.fa' % id, 'w+') as fasta:
for name in names:
seq = name_to_seq[name]
value = gen_test(seq, id, name)
fasta.write(value + '\n')
def main():
if len(sys.argv) != 5:
print(
"Usage: art_test.py grouping.csv scanresults.scan art_out.maf scores.csv"
)
sys.exit(0)
id_to_name = readers.read_grouping(sys.argv[1])
thresh = []
fp = []
tp = []
nf = []
contig_to_read = readers.read_maf(sys.argv[3])
for threshold in range(0, 81, 100):
mismatch = []
scan_id_to_name, scan_name_to_id = readers.read_scan_results(
threshold, sys.argv[2])
scan_name_to_id = substitute_read_name(scan_name_to_id, contig_to_read)
found_score = []
true_positive = 0
false_positive = 0
not_found = 0
total = 0
for id in sorted(id_to_name.keys()):
for name in id_to_name[id]:
total += 1
ids = search(name, id, scan_name_to_id)
if ids:
if ids[0][0] == id or ids[0][0].split('s')[0] == id.split(
's')[0]:
true_positive += 1
found_score.append((ids[0][1], id, name))
else:
#print("False Positive: %s, %s" % (name, id))
canonical_id = 'ARO:' + id.split('s')[0].split('O')[1]
canonical_id2 = 'ARO:' + ids[0][0].split('s')[0].split(
'O')[1]
mismatch.append(
(name, canonical_id, canonical_id2, ids[0][2]))
for i in range(len(ids)):
#print("Attempt %d: %s %f" % (i, ids[i][0], ids[i][1]))
if ids[i][0] == id:
found_score.append((ids[i][1], id, name))
break
false_positive += 1
else:
for k, v in contig_to_read.items():
if name in v:
print("Found read %s in %s" % (name, k))
print("Not Found: %s,%s" % (name, id))
not_found += 1
thresh.append(threshold)
tp.append(true_positive)
fp.append(false_positive)
nf.append(not_found)
print("misclassified %d notfound %d true positive %d of total %d" %
(false_positive, not_found, true_positive, total))
print(threshold, true_positive, false_positive, not_found, total)
x = np.array(fp) # false_positive_rate
y = np.array(tp) # true_positive_rate
t = np.array(thresh)
n = np.array(nf)
found_score.sort()
with open(sys.argv[4], 'w') as f:
for score, hmm, name in found_score:
f.write('%s,%d\n' % (hmm, score - 1))
with open('mismatch.csv', 'w') as f:
f.write('Gene,Expected ID,Found ID\n')
for a, b, c, d in mismatch:
f.write('%s,%s,%s\n' % (b, c, a))
test.graph(x, y, t, n)
def main():
if len(sys.argv) != 4:
print("Usage: test.py grouping.csv scanresults.scan threshold.csv")
sys.exit(0)
id_to_name = readers.read_grouping(sys.argv[1], short=True)
thresh = []
fp = []
tp = []
nf = []
for threshold in range(0, 10, 100):
scan_id_to_name, scan_name_to_id = readers.read_scan_results(threshold, sys.argv[2])
found_score = []
mismatch = []
true_positive = 0
false_positive = 0
not_found = 0
total = 0
for id in id_to_name.keys():
for name in id_to_name[id]:
total += 1
if name not in scan_name_to_id:
print("Not found %s,%s" % (name, id))
not_found += 1
continue
ids = scan_name_to_id[name]
ids.sort(key=lambda l: l[1], reverse=True)
if ids[0][0] == id or ids[0][0].split('s')[0] == id.split('s')[0]:
true_positive += 1
found_score.append((ids[0][1], id, name))
else:
print("False Positive: %s, %s" % (name, id))
if len(ids) > 1 and ids[1][0] == id:
mismatch.append((name, id, ids[0][0]))
else:
for i in range(len(ids)):
print("Attempt %d: %s %f" % (i, ids[i][0], ids[i][1]))
if ids[i][0] == id:
found_score.append((ids[i][1], id, name))
break
else:
mismatch.append((name, id, ids[i][0]))
false_positive += 1
thresh.append(threshold)
tp.append(true_positive)
fp.append(false_positive)
nf.append(not_found)
x = np.array(fp)# false_positive_rate
y = np.array(tp)# true_positive_rate
t = np.array(thresh)
n = np.array(nf)
total = x + y + n
found_score.sort()
with open(sys.argv[3], 'w') as f:
for score, hmm, name in found_score:
f.write('%s,%d\n' % (hmm, score - 1))
with open('mismatch.csv', 'w') as f:
f.write('Gene,Expected ID,Found ID\n')
for a, b, c in mismatch:
f.write('%s,%s,%s\n' % (a, b, c))
print("misclassified %d notfound %d true positive %d of total %d" % (x, n, y, total))
import readers
import sys
# read input files
name_to_sequence = readers.read_fasta(sys.argv[1])
id_to_name = readers.read_grouping(sys.argv[2])
# create fasta files for each id
readers.create_fasta_file_for_each_id(name_to_sequence, id_to_name,
sys.argv[3])