def generate_domain_features(gene: CDSFeature, domains: List[HMMResult]) -> Dict[HMMResult, AntismashDomain]:
""" Generates AntismashDomain features for each provided HMMResult
Arguments:
gene: the CDSFeature the domains were found in
domains: a list of HMMResults found in the CDSFeature
Returns:
a dictionary mapping the HMMResult used to the matching AntismashDomain
"""
new_features = {}
domain_counts = defaultdict(int) # type: Dict[str, int]
for domain in domains:
loc = gene.get_sub_location_from_protein_coordinates(domain.query_start, domain.query_end)
# set up new feature
new_feature = AntismashDomain(loc, tool="nrps_pks_domains")
new_feature.domain = domain.hit_id
new_feature.locus_tag = gene.locus_tag or gene.get_name()
new_feature.detection = "hmmscan"
new_feature.database = "nrpspksdomains.hmm"
new_feature.evalue = domain.evalue
new_feature.score = domain.bitscore
new_feature.translation = gene.translation[domain.query_start:domain.query_end + 1]
domain_counts[domain.hit_id] += 1 # 1-indexed, so increment before use
domain_name = "{}_{}.{}".format(gene.get_name(), domain.hit_id, domain_counts[domain.hit_id])
new_feature.domain_id = "nrpspksdomains_" + domain_name
new_feature.label = domain_name
new_features[domain] = new_feature
return new_features
def generate_domains(self):
inputs = fasta.read_fasta(
path.get_full_path(__file__, 'data', 'PKS_KS.input'))
domains = []
last_end = 0
for translation in inputs.values():
location = FeatureLocation(last_end + 10,
last_end + len(translation) * 3 + 16)
domain = AntismashDomain(location, tool="test")
domain.translation = translation
domains.append(domain)
domain.domain = "PKS_KS"
location = FeatureLocation(
last_end + 10, last_end + len(domains[-1].translation) * 3 + 16)
domains.append(AntismashDomain(location, tool="test"))
domains[-1].domain = "PKS_KR"
return domains