def test_with_secmet(self):
domains = [
SecMetQualifier.Domain("testA", 0.1, 1.1, 3, "test"),
SecMetQualifier.Domain("testB", 5.1, 3.9, 5, "dummy")
]
self.cds.sec_met = SecMetQualifier(domains)
bio = self.convert()
assert "sec_met" not in bio.qualifiers # again, detecting leftover legacy versions
assert len(bio.qualifiers["sec_met_domain"]) == 2
assert bio.qualifiers["sec_met_domain"] == list(map(str, domains))
regen = CDSFeature.from_biopython(bio)
assert regen.sec_met
assert len(regen.sec_met.domains) == len(domains)
assert regen.sec_met.domains == domains
def annotate_orfs(cds_features: List[secmet.CDSFeature], hmm_results: Dict[str, List[HSP]]) -> None:
""" Annotates newly found ORFs with sactipeptide domain information.
This is only relevant for CDS features that did not exist during
the cluster detection stage of antiSMASH.
"""
domains_by_feature: Dict[str, List[SecMetQualifier.Domain]] = defaultdict(list)
for hit_id, results in hmm_results.items():
for result in results:
domain = SecMetQualifier.Domain(result.query_id, result.evalue, result.bitscore, 0, "sactipeptides")
domains_by_feature[hit_id].append(domain)
for cds in cds_features:
domains = domains_by_feature[cds.get_name()]
if domains:
cds.sec_met = SecMetQualifier(domains)
def __init__(self, location: Location, translation: str, locus_tag: str = None,
protein_id: str = None, product: str = "", gene: str = None,
translation_table: int = 1) -> None:
super().__init__(location, feature_type="CDS")
_verify_location(location)
# mandatory
self._gene_functions = GeneFunctionAnnotations()
if not (protein_id or locus_tag or gene):
raise ValueError("CDSFeature requires at least one of: gene, protein_id, locus_tag")
# semi-optional
self.protein_id = _sanitise_id_value(protein_id)
self.locus_tag = _sanitise_id_value(locus_tag)
self.gene = _sanitise_id_value(gene)
self.translation = str(translation)
# optional
if not isinstance(product, str):
raise TypeError("product must be a string, not %s" % type(product))
self.product = product
self.transl_table = int(translation_table)
self._sec_met = SecMetQualifier()
self._nrps_pks = NRPSPKSQualifier(self.location.strand)
self.motifs = [] # type: List[features.CDSMotif]
# runtime-only data
self.region = None # type: Optional[features.Region]
self.unique_id = None # type: Optional[str] # set only when added to a record
def from_biopython(bio_feature: SeqFeature, feature: "CDSFeature" = None, # type: ignore
leftovers: Optional[Dict] = None, record: Any = None) -> "CDSFeature":
if leftovers is None:
leftovers = Feature.make_qualifiers_copy(bio_feature)
# grab mandatory qualifiers
transl_table = 1
if record:
transl_table = record.transl_table
if "transl_table" in leftovers:
transl_table = int(leftovers.pop("transl_table")[0])
# semi-optional qualifiers
protein_id = leftovers.pop("protein_id", [None])[0]
locus_tag = leftovers.pop("locus_tag", [None])[0]
gene = leftovers.pop("gene", [None])[0]
if not (gene or protein_id or locus_tag):
if "pseudo" in leftovers or "pseudogene" in leftovers:
gene = "pseudo%s_%s"
else:
gene = "cds%s_%s"
gene = gene % (bio_feature.location.start, bio_feature.location.end)
name = locus_tag or protein_id or gene
try:
_verify_location(bio_feature.location)
except Exception as err:
message = "invalid location for %s: %s" % (name, str(err))
raise SecmetInvalidInputError(message) from err
try:
translation = _ensure_valid_translation(leftovers.pop("translation", [""])[0],
bio_feature.location, transl_table, record)
except ValueError as err:
raise SecmetInvalidInputError(str(err) + ": %s" % name) from err
feature = CDSFeature(bio_feature.location, translation, gene=gene,
locus_tag=locus_tag, protein_id=protein_id,
translation_table=transl_table)
# grab optional qualifiers
feature.product = leftovers.pop("product", [""])[0]
sec_met = leftovers.pop("sec_met_domain", None)
if sec_met:
feature.sec_met = SecMetQualifier.from_biopython(sec_met)
gene_functions = leftovers.pop("gene_functions", [])
if gene_functions:
feature.gene_functions.add_from_qualifier(gene_functions)
feature.nrps_pks.add_from_qualifier(leftovers.pop("NRPS_PKS", []))
# grab parent optional qualifiers
super(CDSFeature, feature).from_biopython(bio_feature, feature=feature, leftovers=leftovers)
return feature
def annotate(self, tool: str) -> None:
""" Annotates a CDSFeature with the results gathered """
all_matching = set()
if not self.cds.sec_met:
self.cds.sec_met = SecMetQualifier(self.domains)
else:
all_matching.update(set(self.cds.sec_met.domain_ids))
self.cds.sec_met.add_domains(self.domains)
for cluster_type, matching_domains in self.definition_domains.items():
all_matching.update(matching_domains)
for domain in matching_domains:
self.cds.gene_functions.add(GeneFunction.CORE, tool, domain, cluster_type)
# and add all detected domains as ADDITIONAL if not CORE
for secmet_domain in self.cds.sec_met.domains:
if secmet_domain.name in all_matching:
continue
self.cds.gene_functions.add(GeneFunction.ADDITIONAL, secmet_domain.tool,
secmet_domain.name)
def strip_record(record: Record) -> None:
""" Discard antismash specific features and feature qualifiers """
logging.debug(
"Stripping antiSMASH features and annotations from record: %s",
record.id)
record.clear_clusters()
record.clear_superclusters()
record.clear_subregions()
record.clear_regions()
record.clear_antismash_domains()
record.clear_pfam_domains()
# clean up antiSMASH-created CDSMotifs, but leave the rest
motifs = list(record.get_cds_motifs())
record.clear_cds_motifs()
for motif in motifs:
if not motif.created_by_antismash:
record.add_cds_motif(motif)
# clean up antiSMASH annotations in CDS features
for feature in record.get_cds_features():
feature.sec_met = SecMetQualifier()
feature.gene_functions.clear()
def test_with_no_secmet(self):
cds = self.create_cds(55000, 60000, profiles=[])
cds.sec_met = SecMetQualifier()
self.record.add_cds_feature(cds)
assert utils.distance_to_pfam(self.record, self.query, ["test"]) == -1
def get_domains_for_cds(cds: CDSFeature) -> List[SecMetQualifier.Domain]:
domains = []
for hsp in results_by_id.get(cds.get_name(), []):
domains.append(SecMetQualifier.Domain(hsp.query_id, hsp.evalue, hsp.bitscore,
num_seeds_per_hmm[hsp.query_id], tool))
return domains
def strip_antismash_annotations(self) -> None:
""" Remove all antiSMASH-specific annotations from the feature """
self.sec_met = SecMetQualifier()
self.gene_functions.clear()
self.nrps_pks = NRPSPKSQualifier(self.location.strand)
def from_biopython(bio_feature: SeqFeature, feature: "CDSFeature" = None, # type: ignore
leftovers: Optional[Dict] = None, record: Any = None) -> "CDSFeature":
if leftovers is None:
leftovers = Feature.make_qualifiers_copy(bio_feature)
# grab mandatory qualifiers
transl_table = 1
if record:
transl_table = record.transl_table
if "transl_table" in leftovers:
transl_table = int(leftovers.pop("transl_table")[0])
translation = leftovers.pop("translation", [""])[0]
# semi-optional qualifiers
protein_id = leftovers.pop("protein_id", [None])[0]
locus_tag = leftovers.pop("locus_tag", [None])[0]
gene = leftovers.pop("gene", [None])[0]
if not (gene or protein_id or locus_tag):
if "pseudo" in leftovers or "pseudogene" in leftovers:
gene = "pseudo%s_%s"
else:
gene = "cds%s_%s"
gene = gene % (bio_feature.location.start, bio_feature.location.end)
try:
_verify_location(bio_feature.location)
except Exception as err:
message = "invalid location for %s: %s" % (gene or protein_id or locus_tag, str(err))
raise SecmetInvalidInputError(message) from err
# ensure translation is valid if it exists
if translation:
invalid = set(translation) - _VALID_TRANSLATION_CHARS
if invalid:
logging.warning("Regenerating translation for CDS %s (at %s) containing invalid characters: %s",
locus_tag or protein_id or gene, bio_feature.location, invalid)
translation = ""
# ensure that the translation fits
if not _is_valid_translation_length(translation, bio_feature.location):
raise SecmetInvalidInputError("translation longer than location allows: %s > %s" % (
len(translation) * 3, len(bio_feature.location)))
# finally, generate the translation if it doesn't exist
if not translation:
if not record:
raise SecmetInvalidInputError("no translation in CDS and no record to generate it with")
if bio_feature.location.end > len(record.seq):
raise SecmetInvalidInputError("feature missing translation and sequence too short: %s" % (
(gene or protein_id or locus_tag)))
translation = record.get_aa_translation_from_location(bio_feature.location, transl_table)
assert _is_valid_translation_length(translation, bio_feature.location)
feature = CDSFeature(bio_feature.location, translation, gene=gene,
locus_tag=locus_tag, protein_id=protein_id,
translation_table=transl_table)
# grab optional qualifiers
feature.product = leftovers.pop("product", [""])[0]
sec_met = leftovers.pop("sec_met_domain", None)
if sec_met:
feature.sec_met = SecMetQualifier.from_biopython(sec_met)
gene_functions = leftovers.pop("gene_functions", [])
if gene_functions:
feature.gene_functions.add_from_qualifier(gene_functions)
# grab parent optional qualifiers
super(CDSFeature, feature).from_biopython(bio_feature, feature=feature, leftovers=leftovers)
return feature
def detect_protoclusters_and_signatures(record: Record, signature_file: str,
seeds_file: str, rule_files: List[str],
filter_file: str,
tool: str) -> RuleDetectionResults:
""" Compares all CDS features in a record with HMM signatures and generates
Protocluster features based on those hits and the current protocluster detection
rules.
Arguments:
record: the record to analyse
signature_file: a tab separated file; each row being a single HMM reference
with columns: label, description, minimum score cutoff, hmm path
seeds_file: the file containing all HMM profiles
rule_files: the files containing the rules to use for cluster definition
filter_file: a file containing equivalence sets of HMMs
tool: the name of the tool providing the HMMs (e.g. clusterfinder, rule_based_clusters)
"""
if not rule_files:
raise ValueError("rules must be provided")
full_fasta = fasta.get_fasta_from_record(record)
# if there's no CDS features, don't try to do anything
if not full_fasta:
return RuleDetectionResults({}, tool)
sig_by_name = {
sig.name: sig
for sig in get_signature_profiles(signature_file)
}
rules = [] # type: List[rule_parser.DetectionRule]
for rule_file in rule_files:
rules = create_rules(rule_file, set(sig_by_name), rules)
results = []
results_by_id = {} # type: Dict[str, HSP]
runresults = run_hmmsearch(seeds_file, full_fasta, use_tempfile=True)
for runresult in runresults:
acc = runresult.accession.split('.')[0]
# Store result if it is above cut-off
for hsp in runresult.hsps:
if hsp.query_id in sig_by_name:
sig = sig_by_name[hsp.query_id]
elif acc in sig_by_name:
sig = sig_by_name[acc]
else:
raise ValueError(
'Failed to find signature for ID %s / ACC %s' %
(hsp.query_id, acc))
if hsp.bitscore > sig.cutoff:
results.append(hsp)
if hsp.hit_id not in results_by_id:
results_by_id[hsp.hit_id] = [hsp]
else:
results_by_id[hsp.hit_id].append(hsp)
# Filter results by comparing scores of different models (for PKS systems)
results, results_by_id = filter_results(results, results_by_id,
filter_file, set(sig_by_name))
# Filter multiple results of the same model in one gene
results, results_by_id = filter_result_multiple(results, results_by_id)
# Use rules to determine gene clusters
cds_domains_by_cluster, cluster_type_hits = apply_cluster_rules(
record, results_by_id, rules)
# Find number of sequences on which each pHMM is based
num_seeds_per_hmm = get_sequence_counts(signature_file)
# Save final results to record
rules_by_name = {rule.name: rule for rule in rules}
clusters = find_protoclusters(record, cluster_type_hits, rules_by_name)
strip_inferior_domains(cds_domains_by_cluster, rules_by_name)
cds_results_by_cluster = {}
for cluster in clusters:
cds_results = []
for cds in record.get_cds_features_within_location(cluster.location):
domains = []
for hsp in results_by_id.get(cds.get_name(), []):
domains.append(
SecMetQualifier.Domain(hsp.query_id, hsp.evalue,
hsp.bitscore,
num_seeds_per_hmm[hsp.query_id],
tool))
if domains:
cds_results.append(
CDSResults(cds, domains,
cds_domains_by_cluster.get(cds.get_name(), {})))
cds_results_by_cluster[cluster] = cds_results
return RuleDetectionResults(cds_results_by_cluster, tool)