def main():
threshold = 150
scan_id_to_name, scan_name_to_id = readers.read_scan_results(threshold, sys.argv[1])
genes = [(len(v), k) for k,v in scan_id_to_name.items()]
genes.sort(reverse=True)
for c,g in genes:
print('%s,%d' % (g, c))
def main():
threshold = 150
scan_id_to_name, scan_name_to_id = readers.read_scan_results(
threshold, sys.argv[1])
genes = [(len(v), k) for k, v in scan_id_to_name.items()]
genes.sort(reverse=True)
for c, g in genes:
print('%s,%d' % (g, c))
def main():
if len(sys.argv) < 3 or len(sys.argv) > 4:
print("Usage: functional_test.py [-protein] results.scan seq.fasta")
sys.exit(-1)
protein = True if sys.argv[1] == '-protein' else False
if protein:
scan_file = sys.argv[2]
fasta_file = sys.argv[3]
protein_map = {
} # place to store mapping of protein name to dna name for genes
else:
scan_file = sys.argv[1]
fasta_file = sys.argv[2]
genes = readers.read_fasta(fasta_file)
positive_count = 0
negative_count = 0
for gene in genes.keys():
if 'True' in gene:
positive_count += 1
elif 'False' in gene:
negative_count += 1
if protein:
names = gene.split('>')
protein_map[names[0].strip()] = names[1].strip()
id_to_target, target_to_id = readers.read_scan_results(0,
scan_file,
protein=protein)
false_positive = 0
true_positive = 0
already_seen_protein = set(
) # We don't want to double count if we have seen the same gene
for key in target_to_id.keys():
if protein:
key = protein_map[key]
if key in already_seen_protein:
continue
else:
already_seen_protein.add(key)
if key.startswith('False'):
false_positive += 1
elif key.startswith('True'):
true_positive += 1
print("True Positive: %d/%d(%f); False Positive: %d/%d(%f)" %
(true_positive, positive_count,
float(true_positive) / positive_count, false_positive,
negative_count, float(false_positive) / negative_count))
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)
def main():
if len(sys.argv) < 3 or len(sys.argv) > 4:
print(
"Usage: functional_compare.py [-fsl] [-protein] results.scan seq.fasta"
)
sys.exit(-1)
fsl = True if sys.argv[1] == '-fsl' else False
protein = True if sys.argv[1] == '-protein' else False
if fsl:
fsl_file = sys.argv[2]
fasta_file = sys.argv[3]
elif protein:
scan_file = sys.argv[2]
fasta_file = sys.argv[3]
else:
scan_file = sys.argv[1]
fasta_file = sys.argv[2]
genes = readers.read_fasta(fasta_file)
if fsl:
target_to_id = read_fsl(fsl_file)
else:
id_to_target, target_to_id = readers.read_scan_results(0,
scan_file,
protein=protein)
if protein:
readers.change_RF_to_ARO(target_to_id)
already_seen_protein = set(
) # We don't want to double count if we have seen the same gene
terms = ontology_common.parse_obo('new_combined.obo')
false_positive = 0
true_positive = 0
false_negative = 0
for gene in genes.keys():
if protein:
names = gene.split('>')
gene = names[1].strip()
name = names[0].strip()
if gene in already_seen_protein:
continue
else:
already_seen_protein.add(gene)
else:
name = gene
found = name in target_to_id
if found:
antibiotic = gene.split('_')[1]
functional_antibiotic = antibiotic_code[antibiotic]
results = target_to_id[name]
results.sort(key=lambda l: l[1], reverse=True)
index = 0
while index < len(results):
result = results[index]
index += 1
id = result[0]
# remove formatting used by hmm
if 's' in id:
id = id.replace('ARO', 'ARO:')
id = id.split('s')[0]
if ';' in id:
# resfams can have a list of ids associated with a gene
classes = [
terms[p]['name'] for i in id.split(';')
for p in ontology_common.get_class(i, terms)
]
drugs = set()
for i in id.split(';'):
drugs |= ontology_common.get_resistance(
ontology_common.get_lineage(i, terms), terms)
else:
classes = [
terms[p]['name']
for p in ontology_common.get_class(id, terms)
]
drugs = ontology_common.get_resistance(
ontology_common.get_lineage(id, terms), terms)
identified = False
for drug in drugs:
for d in ontology_common.get_lineage(drug, terms):
for fd in ontology_common.get_lineage(
functional_antibiotic[1], terms):
if d == fd and d not in [
'ARO:1000001', 'ARO:1000003', 'Unknown'
]:
identified = True
if identified:
true_positive += 1
break
else:
false_negative += 1
if found and not identified:
print(gene, functional_antibiotic, id, classes, drugs)
false_positive += 1
print('False negative: %d; False Positive:%d; True Positive:%d' %
(false_negative, false_positive, true_positive))
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) < 3 or len(sys.argv) > 4:
print("Usage: functional_compare.py [-fsl] [-protein] results.scan seq.fasta")
sys.exit(-1)
fsl = True if sys.argv[1] == '-fsl' else False
protein = True if sys.argv[1] == '-protein' else False
if fsl:
fsl_file = sys.argv[2]
fasta_file = sys.argv[3]
elif protein:
scan_file = sys.argv[2]
fasta_file = sys.argv[3]
else:
scan_file = sys.argv[1]
fasta_file = sys.argv[2]
genes = readers.read_fasta(fasta_file)
if fsl:
target_to_id = read_fsl(fsl_file)
else:
id_to_target, target_to_id = readers.read_scan_results(0, scan_file, protein=protein)
if protein:
readers.change_RF_to_ARO(target_to_id)
already_seen_protein = set() # We don't want to double count if we have seen the same gene
terms = ontology_common.parse_obo('new_combined.obo')
false_positive = 0
true_positive = 0
false_negative = 0
for gene in genes.keys():
if protein:
names = gene.split('>')
gene = names[1].strip()
name = names[0].strip()
if gene in already_seen_protein:
continue
else:
already_seen_protein.add(gene)
else:
name = gene
found = name in target_to_id
if found:
antibiotic = gene.split('_')[1]
functional_antibiotic = antibiotic_code[antibiotic]
results = target_to_id[name]
results.sort(key=lambda l: l[1], reverse=True)
index = 0
while index < len(results):
result = results[index]
index += 1
id = result[0]
# remove formatting used by hmm
if 's' in id:
id = id.replace('ARO', 'ARO:')
id = id.split('s')[0]
if ';' in id:
# resfams can have a list of ids associated with a gene
classes = [terms[p]['name'] for i in id.split(';') for p in ontology_common.get_class(i, terms)]
drugs = set()
for i in id.split(';'):
drugs |= ontology_common.get_resistance(ontology_common.get_lineage(i, terms), terms)
else:
classes = [terms[p]['name'] for p in ontology_common.get_class(id, terms)]
drugs = ontology_common.get_resistance(ontology_common.get_lineage(id, terms), terms)
identified = False
for drug in drugs:
for d in ontology_common.get_lineage(drug, terms):
for fd in ontology_common.get_lineage(functional_antibiotic[1], terms):
if d == fd and d not in ['ARO:1000001', 'ARO:1000003', 'Unknown']:
identified = True
if identified:
true_positive += 1
break
else:
false_negative += 1
if found and not identified:
print(gene, functional_antibiotic, id, classes, drugs)
false_positive += 1
print('False negative: %d; False Positive:%d; True Positive:%d' % (false_negative, false_positive, true_positive))
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))