def scan_for_functions(cds_features: List[CDSFeature], database: str,
hmmscan_opts: Optional[List[str]] = None) -> Dict[str, HMMResult]:
""" Finds possible classifications for the provided genes.
Arguments:
cds_features: a list of CDSFeatures to classify
database: the path to the database to check
hmmscan_opts: a list of extra options to provide to hmmscan
Returns:
a dictionary mapping CDS name to a list of HMMResult instances of
classifications
"""
search_fasta = fasta.get_fasta_from_features(cds_features)
results = subprocessing.run_hmmscan(database, search_fasta, hmmscan_opts)
hmm_lengths = utils.get_hmm_lengths(database)
hmm_results = refine_hmmscan_results(results, hmm_lengths)
best_hits = {} # type: Dict[str, HMMResult]
for cds in cds_features:
cds_name = cds.get_name()
hits = hmm_results.get(cds_name)
if not hits:
continue
best_hits[cds_name] = hits[0]
return best_hits
def run_hmmer(record: Record, features: Iterable[CDSFeature], max_evalue: float,
min_score: float, database: str, tool: str, filter_overlapping: bool = True
) -> HmmerResults:
""" Build hmmer results for the given features
Arguments:
record: the Record instance to run hmmer over
features: the list of CDSFeatures to run over specifically
max_evalue: a maximum evalue allowed for hits (exclusive)
min_evalue: a minimum evalue allowed for hits (exclusive)
database: the database to search for hits within
tool: the name of the specific tool calling into this module
"""
if not os.path.exists(database):
raise ValueError("Given database does not exist: %s" % database)
query_sequence = fasta.get_fasta_from_features(features)
hmmscan_results = subprocessing.run_hmmscan(database, query_sequence, opts=["--cut_tc"])
hits = build_hits(record, hmmscan_results, min_score, max_evalue, database)
if filter_overlapping:
results_by_cds = defaultdict(list)
for hit in hits:
results_by_cds[hit.locus_tag].append(hit)
cutoffs = pfamdb.get_pfam_cutoffs(database)
hits = []
for locus_hits in results_by_cds.values():
hits.extend(remove_overlapping(locus_hits, cutoffs))
return HmmerResults(record.id, max_evalue, min_score, database, tool, hits)
def run_hmmer(record: Record, features: List[CDSFeature], max_evalue: float,
min_score: float, database: str, tool: str) -> HmmerResults:
""" Build hmmer results for the given features"""
if not os.path.exists(database):
raise ValueError("Given database does not exist: %s" % database)
query_sequence = fasta.get_fasta_from_features(features)
hmmscan_results = subprocessing.run_hmmscan(database, query_sequence)
hits = build_hits(record, hmmscan_results, min_score, max_evalue, database)
return HmmerResults(record.id, max_evalue, min_score, database, tool, hits)
def find_ks_domains(fasta: str) -> Dict[str, List[HMMResult]]:
""" Analyse KS domains & PKS/NRPS protein domain composition to detect NRPS/PKS types
Arguments:
fasta: a group of features in fasta format
Returns:
a dictionary mapping feature name to a list of KS domain results for that feature
"""
opts = ["--cut_tc"]
ks_file = path.get_full_path(__file__, "data", "ksdomains.hmm")
lengths = utils.get_hmm_lengths(ks_file)
domains = subprocessing.run_hmmscan(ks_file, fasta, opts)
return refine_hmmscan_results(domains, lengths, neighbour_mode=True)
def find_ab_motifs(fasta: str) -> Dict[str, List[HMMResult]]:
""" Analyse for abMotifs
Arguments:
fasta: a group of features in fasta format
Returns:
a dictionary mapping feature name to a list of motif results for that feature
"""
opts = ["-E", "0.25"]
motif_file = path.get_full_path(__file__, "data", "abmotifs.hmm")
abmotif_results = subprocessing.run_hmmscan(motif_file, fasta, opts)
lengths = utils.get_hmm_lengths(motif_file)
return refine_hmmscan_results(abmotif_results, lengths, neighbour_mode=True)
def run_t2pks_hmmscan(cluster: Cluster) -> Dict[str, List[HMMResult]]:
""" Runs hmmscan for type II PKS proteins on coding sequences in cluster
Arguments:
cluster: Cluster on which the type II PKS hmmscan shall be run
Returns:
a dictionary of key: cds and value: list of HMMResults, for hmmscan results of the cluster
"""
cluster_fasta = fasta.get_fasta_from_features(cluster.cds_children)
hmm_file = path.get_full_path(__file__, "data", "t2pks.hmm")
hmm_results = subprocessing.run_hmmscan(hmm_file,
cluster_fasta,
opts=['--cut_tc'])
hmm_lengths = get_hmm_lengths(hmm_file)
return refine_hmmscan_results(hmm_results, hmm_lengths)
def find_domains(fasta: str, record: Record) -> Dict[str, List[HMMResult]]:
""" Analyse for C/A/PCP/E/KS/AT/ATd/DH/KR/ER/ACP/TE/TD/COM/Docking/MT/CAL domains
Arguments:
fasta: a group of features in fasta format
record: the Record that contains all the features
Returns:
a dictionary mapping feature name to a list of domain results for that feature
"""
opts = ["--cut_tc"]
nrpspks_file = path.get_full_path(__file__, "data", "nrpspksdomains.hmm")
nrpspksdomain_results = subprocessing.run_hmmscan(nrpspks_file, fasta, opts)
lengths = utils.get_hmm_lengths(nrpspks_file)
domains = refine_hmmscan_results(nrpspksdomain_results, lengths, neighbour_mode=True)
return filter_nonterminal_docking_domains(record, domains)
def run_t2pks_hmmscan(
cds_features: Iterable[CDSFeature]) -> Dict[str, List[HMMResult]]:
""" Runs hmmscan for type II PKS proteins on the given CDSFeatures
Arguments:
cluster: Protocluster on which to run the type II PKS hmmscan
Returns:
a dictionary of key: cds and value: list of HMMResults, for hmmscan results of the cluster
"""
cluster_fasta = fasta.get_fasta_from_features(cds_features)
hmm_file = path.get_full_path(__file__, "data", "t2pks.hmm")
hmm_results = subprocessing.run_hmmscan(hmm_file,
cluster_fasta,
opts=['--cut_tc'])
hmm_lengths = get_hmm_lengths(hmm_file)
return refine_hmmscan_results(hmm_results, hmm_lengths)
def run_hmmer(record: Record, features: Iterable[CDSFeature], max_evalue: float,
min_score: float, database: str, tool: str) -> HmmerResults:
""" Build hmmer results for the given features
Arguments:
record: the Record instance to run hmmer over
features: the list of CDSFeatures to run over specifically
max_evalue: a maximum evalue allowed for hits (exclusive)
min_evalue: a minimum evalue allowed for hits (exclusive)
database: the database to search for hits within
tool: the name of the specific tool calling into this module
"""
if not os.path.exists(database):
raise ValueError("Given database does not exist: %s" % database)
query_sequence = fasta.get_fasta_from_features(features)
hmmscan_results = subprocessing.run_hmmscan(database, query_sequence, opts=["--cut_tc"])
hits = build_hits(record, hmmscan_results, min_score, max_evalue, database)
return HmmerResults(record.id, max_evalue, min_score, database, tool, hits)
def classify_genes(
cds_features: List[CDSFeature]) -> Dict[str, List[HMMResult]]:
""" Finds possible classifications for the provided genes.
Arguments:
cds_features: a list of CDSFeatures to classify
Returns:
a dictionary mapping CDS name to a list of HMMResult instances of
classifications
"""
smcogs_fasta = fasta.get_fasta_from_features(cds_features)
smcogs_opts = ["-E", "1E-6"]
hmm_file = path.get_full_path(__file__, "data", "smcogs.hmm")
smcogs_results = subprocessing.run_hmmscan(hmm_file, smcogs_fasta,
smcogs_opts)
hmm_lengths = utils.get_hmm_lengths(hmm_file)
return refine_hmmscan_results(smcogs_results, hmm_lengths)
