def get_34_aa_signature(domain: AntismashDomain) -> str:
""" Extract 10 / 34 AA NRPS signatures from A domains """
assert " " not in domain.get_name()
assert verify_good_sequence(domain.translation)
# Run muscle and collect sequence positions from file
alignments = subprocessing.run_muscle_single(domain.get_name(),
domain.translation,
ADOMAINS_FILENAME)
domain_alignment = alignments[domain.get_name()]
reference_alignment = alignments[REF_SEQUENCE]
positions = read_positions(APOSITION_FILENAME, START_POSITION)
# Count residues in ref sequence and put positions in list
poslist = build_position_list(positions, reference_alignment)
# Extract positions from query sequence
query_sig_seq = extract(domain_alignment, poslist)
# Add fixed lysine 517
query_sig_seq += "K"
# repeat with 34 AA codes
angpositions = read_positions(A34_POSITIONS_FILENAME, START_POSITION)
poslist = build_position_list(angpositions, reference_alignment)
return extract(domain_alignment, poslist)
def test_angstrom(self):
domain = AntismashDomain(FeatureLocation(1, 2), "test")
domain.domain_id = "query"
domain.translation = self.aligns[domain.domain_id].replace("-", "")
sig = nrps_predictor.get_34_aa_signature(domain)
assert sig == "L--SFDASLFEMYLLTGGDRNMYGPTEATMCATW"
def main():
domain = AntismashDomain(FeatureLocation(1, 3, 1), tool="test")
print(get_34_aa_signature(domain))
out_file = fasta_dir
# Reads in file
try:
create_domain_fa = fasta.read_fasta(out_file)
except:
print("Error: please check your file is a valid FASTA or GenBank file")
sys.exit(1)
# Create antiSMASH-like domain objects from AMP-binding hits
domain_list = []
for i, domain in enumerate(create_domain_fa):
domain_list.append(
AntismashDomain(
FeatureLocation(1, 1, 1),
tool="test")) # arbitrary feature location for testing
domain_list[i].domain_id = list(create_domain_fa.keys())[i]
domain_list[i].translation = list(create_domain_fa.values())[i]
# Extract the active site residues
res, nms, sqs = [], [], []
new_path = "%s/data/34_aa_xtracted.fasta" % parent_folder
if not silent:
print("##### Extracting active site residues... #####")
for i, domain in enumerate(tqdm.tqdm(domain_list, disable=silent)):
res.append(get_34_aa_signature(domain))
nms.append(domain.domain_id)
sqs.append(res[i])
def add_to_record(self, record) -> None:
""" Save substrate specificity predictions in NRPS/PKS domain sec_met info of record
"""
for feature in record.get_nrps_pks_cds_features():
x_count = 0
nrps_qualifier = feature.nrps_pks
new_features = []
gene_id = feature.get_name()
for domain in nrps_qualifier.domains:
domain_type = domain.name
start_aa = domain.start
end_aa = domain.end
evalue = domain.evalue
score = domain.bitscore
domain.predictions.clear()
# calculate respective positions based on aa coordinates
if feature.location.strand == 1:
start = feature.location.start + (3 * start_aa)
end = feature.location.start + (3 * end_aa)
else:
end = feature.location.end - (3 * start_aa)
start = feature.location.end - (3 * end_aa)
loc = FeatureLocation(start, end, strand=feature.strand)
# set up new CDS_motif feature
new_feature = AntismashDomain(loc)
new_feature.domain_subtype = domain_type
if feature.locus_tag:
new_feature.locus_tag = feature.locus_tag
else:
new_feature.locus_tag = gene_id
new_feature.detection = "hmmscan"
new_feature.database = "nrpspksdomains.hmm"
new_feature.evalue = evalue
new_feature.score = score
if feature.transl_table:
transl_table = feature.transl_table
else:
transl_table = 1
new_feature.translation = str(new_feature.extract(record.seq).translate(table=transl_table))
domainname = gene_id + domain.label
if domain_type == "AMP-binding":
new_feature.label = domainname
new_feature.domain_id = "nrpspksdomains_" + domainname
domain.predictions["consensus"] = "nrp"
elif domain_type == "PKS_AT":
new_feature.label = domainname
new_feature.domain_id = "nrpspksdomains_" + domainname
# For t1pks, t2pks and t3pks
if 'transatpks' not in feature.cluster.products:
consensus = self.consensus[domainname]
else: # for transatpks
consensus = self.consensus_transat[domainname]
pks_sig = self.pks.method_results["signature"][domainname]
if pks_sig:
domain.predictions["PKS signature"] = pks_sig[0].name.rsplit("_", 1)[1]
else:
domain.predictions["PKS signature"] = _UNKNOWN
minowa = self.pks.method_results["minowa_at"][domainname][0][0]
domain.predictions["Minowa"] = LONG_TO_SHORT.get(minowa, minowa)
domain.predictions["consensus"] = consensus
elif domain_type == "CAL_domain":
new_feature.label = domainname
new_feature.domain_id = "nrpspksdomains_" + domainname
minowa = self.pks.method_results["minowa_cal"][domainname][0][0]
domain.predictions["Minowa"] = LONG_TO_SHORT.get(minowa, minowa)
elif domain_type == "PKS_KR":
new_feature.label = domainname
new_feature.domain_id = "nrpspksdomains_" + domainname
domain.predictions["KR activity"] = \
"active" if self.pks.method_results["kr_activity"][domainname] else "inactive"
domain.predictions["KR stereochemistry"] = \
self.pks.method_results["kr_stereochem"].get(domainname, _UNKNOWN)
else:
x_count += 1
new_feature.domain_id = "nrpspksdomains_" + gene_id.partition(".")[0] \
+ "_Xdom"+'{:02d}'.format(x_count)
# updated_nrps_qualifier.append(domain) # TODO weird, but should it be done?
for method, pred in domain.predictions.items():
new_feature.specificity.append("%s: %s" % (method, pred))
mapping = DOMAIN_TYPE_MAPPING.get(domain_type)
if mapping:
new_feature.domain_subtype = domain_type
new_feature.domain = mapping
new_features.append(new_feature)
for new_feature in new_features:
record.add_feature(new_feature)
with open(input_filename, "w") as handle:
for sig, domain in zip(signatures, a_domains):
handle.write("%s\t%s\n" % (sig, domain.get_name()))
# Run NRPSPredictor2 SVM
commands = [
'java',
'-Ddatadir=%s' % data_dir, '-cp', classpath,
'org.roettig.NRPSpredictor2.NRPSpredictor2', '-i', input_filename,
'-r', output_filename, '-s', '1', '-b', bacterial
]
result = subprocessing.execute(commands)
if not result.successful():
raise RuntimeError("NRPSPredictor2 failed: %s" % result.stderr)
with open(output_filename) as handle:
lines = handle.read().splitlines()[1:] # strip the header
return read_output(lines)
create_domain_fa = fasta.read_fasta(
'/Users/robi0916/Documents/Wageningen_UR/github/sandpuma2_serina/flat/fullset20160624_cl.faa'
)
domain_list = []
for i, domain in enumerate(create_domain_fa):
domain_list.append(AntismashDomain(FeatureLocation(
1, 1, 1), tool="test")) # arbitrary feature location
domain_list[i].domain_id = list(create_domain_fa.keys())[i]
domain_list[i].translation = list(create_domain_fa.values())[i]
run_nrpspredictor(domain_list)