def test_limit_to_record_complete(self):
records = self.read_double_nisin()
config.update_config({"limit_to_record": "bad_id"})
with self.assertRaisesRegex(AntismashInputError,
"no sequences matched filter"):
record_processing.pre_process_sequences(records, self.options,
self.genefinding)
def test_nisin_fasta_only(self):
config.update_config({"genefinding_tool": "none"})
filepath = path.get_full_path(__file__, "data", "nisin.fasta")
records = record_processing.parse_input_sequence(filepath)
assert len(records) == 1
assert not records[0].get_cds_features()
# make sure genefinding wasn't run with default options
with self.assertRaisesRegex(AntismashInputError,
"all records skipped"):
record_processing.pre_process_sequences(records, self.options,
self.genefinding)
assert not self.genefinding.was_run
assert not records[0].get_cds_features()
# make sure genefinding was run when not 'none'
records[0].skip = False
config.update_config({"genefinding_tool": "not-none"})
# due to no genes actually being marked, it'll raise an error
with self.assertRaisesRegex(AntismashInputError,
"all records skipped"):
record_processing.pre_process_sequences(records, self.options,
self.genefinding)
# but genefinding was still run
assert self.genefinding.was_run
# still no features because we used dummy genefinding
for record in records:
assert not record.get_cds_features()
assert record.skip.lower() == "no genes found"
def test_shotgun(self):
filepath = path.get_full_path(__file__, "data", "wgs.gbk")
records = record_processing.parse_input_sequence(filepath)
with self.assertRaisesRegex(
AntismashInputError,
"incomplete whole genome shotgun records are not supported"):
record_processing.pre_process_sequences(records, self.options,
self.genefinding)
def test_limit_to_record_partial(self):
records = self.read_double_nisin()
assert all(rec.skip is None for rec in records)
config.update_config({"limit_to_record": records[0].id})
records[0].id += "_changed"
record_processing.pre_process_sequences(records, self.options, self.genefinding)
assert not records[1].skip
assert records[0].skip.startswith("did not match filter")
def test_nisin_fasta_gff(self):
fasta = path.get_full_path(__file__, "data", "nisin.fasta")
gff = path.get_full_path(__file__, "data", "nisin.gff3")
config.update_config({"genefinding_gff3": gff})
records = record_processing.parse_input_sequence(fasta, gff_file=gff)
record_processing.pre_process_sequences(records, self.options, self.genefinding)
assert not self.genefinding.was_run
assert len(records[0].get_cds_features()) == 11
def test_limit(self):
records = self.read_double_nisin()
assert all(rec.skip is None for rec in records)
assert not self.options.triggered_limit
config.update_config({"limit": 1})
record_processing.pre_process_sequences(records, self.options, self.genefinding)
assert records[0].skip is None
assert records[1].skip.startswith("skipping all but first 1")
assert self.options.triggered_limit
def test_duplicate_record_ids(self):
records = self.read_double_nisin()
assert records[0].id == records[1].id
records = record_processing.pre_process_sequences(
records, self.options, self.genefinding)
assert len(records) == 2
assert records[0].id != records[1].id
def test_nisin(self):
record = parse_input_sequence(get_path_to_nisin_fasta())[0]
assert record.get_feature_count() == 0
record = pre_process_sequences([record], self.options, genefinding)[0]
assert record.get_feature_count() == 12
# and make sure they're all CDS features
assert len(record.get_cds_features()) == 12
def test_fumigatus_cluster(self):
record = parse_input_sequence(self.data_file('fumigatus.cluster1.fna'),
taxon="fungi")[0]
assert record.get_feature_count() == 0
record = pre_process_sequences([record], self.options, genefinding)[0]
assert record.get_feature_count() == 11
# and make sure they're all CDS features
assert len(record.get_cds_features()) == 11
def test_nisin_fasta_only(self):
config.update_config({"genefinding_tool": "none"})
filepath = path.get_full_path(__file__, "data", "nisin.fasta")
records = record_processing.parse_input_sequence(filepath)
assert len(records) == 1
assert not records[0].get_cds_features()
# make sure genefinding wasn't run with default options
record_processing.pre_process_sequences(records, self.options,
self.genefinding)
assert not self.genefinding.was_run
assert not records[0].get_cds_features()
# make sure genefinding was run when not 'none'
records[0].skip = False
config.update_config({"genefinding_tool": "not-none"})
record_processing.pre_process_sequences(records, self.options,
self.genefinding)
assert self.genefinding.was_run
# still no features because we used dummy genefinding
assert not records[0].get_cds_features()
def test_records_with_bad_names(self):
# reuse fumigatus and change the id to bad ids
for bad in [
".", # changes due to glimmerhmm
"-bad", # could cause a fasta file to be created that is interpreted as an arg
]:
record = parse_input_sequence(
self.data_file('fumigatus.cluster1.fna'), taxon="fungi")[0]
record.id = bad
record = pre_process_sequences([record], self.options,
genefinding)[0]
assert record.get_cds_features()
def _run_antismash(sequence_file: Optional[str], options: ConfigType) -> int:
""" The real run_antismash, assumes logging is set up around it """
logging.info("antiSMASH version: %s", options.version)
_log_found_executables(options)
detection_modules = get_detection_modules()
analysis_modules = get_analysis_modules()
output_modules = get_output_modules()
modules = detection_modules + analysis_modules + output_modules
if options.list_plugins:
list_plugins(modules)
return 0
options.all_enabled_modules = list(filter(lambda x: x.is_enabled(options), modules))
if options.check_prereqs_only:
try:
check_prerequisites(modules, options)
except RuntimeError:
print("Some module prerequisites not satisfied")
return 1
print("All prerequisites satisfied")
return 0
else:
check_prerequisites(options.all_enabled_modules, options)
# start up profiling if relevant
if options.profile:
profiler = cProfile.Profile()
profiler.enable()
# ensure the provided options are valid
if not verify_options(options, options.all_enabled_modules):
return 1
# check that at least one module will run
if not options.all_enabled_modules:
raise ValueError("No detection or analysis modules enabled")
start_time = datetime.now()
results = read_data(sequence_file, options)
# reset module timings
results.timings_by_record.clear()
prepare_output_directory(options.output_dir, sequence_file or options.reuse_results)
results.records = record_processing.pre_process_sequences(results.records, options,
cast(AntismashModule, genefinding))
for record, module_results in zip(results.records, results.results):
# skip if we're not interested in it
if record.skip:
continue
logging.info("Analysing record: %s", record.id)
timings = run_detection(record, options, module_results)
# and skip analysis if detection didn't find anything
if not record.get_regions():
continue
analysis_timings = analyse_record(record, options, analysis_modules, module_results)
timings.update(analysis_timings)
results.timings_by_record[record.id] = timings
# Write results
json_filename = os.path.join(options.output_dir, results.input_file)
json_filename = os.path.splitext(json_filename)[0] + ".json"
logging.debug("Writing json results to '%s'", json_filename)
results.write_to_file(json_filename)
# now that the json is out of the way, annotate the record
# otherwise we could double annotate some areas
annotate_records(results)
# create relevant output files
write_outputs(results, options)
# save profiling data
if options.profile:
profiler.disable()
write_profiling_results(profiler, os.path.join(options.output_dir,
"profiling_results"))
running_time = datetime.now() - start_time
# display module runtimes before total time
if options.debug:
log_module_runtimes(results.timings_by_record)
logging.debug("antiSMASH calculation finished at %s; runtime: %s",
datetime.now().strftime("%Y-%m-%d %H:%M:%S"), str(running_time))
logging.info("antiSMASH status: SUCCESS")
return 0
def run_antismash(sequence_file: Optional[str], options) -> int:
""" The complete antismash pipeline. Reads in data, runs detection and
analysis modules over any records found, then outputs the results to
file.
Arguments:
sequence_file: the sequence file to read in records from, can be
None if reusing results
options: command line options as an argparse.Namespace
detection_modules: None or a list of modules to use for detection,
if None defaults will be used
analysis_modules: None or a list of modules to use for analysis,
if None defaults will be used
Returns:
0 if requested operations completed succesfully, otherwise 1
Exceptions may also be raised
"""
logfile = options.logfile
setup_logging(logfile=logfile,
verbose=options.verbose,
debug=options.debug)
detection_modules = get_detection_modules()
analysis_modules = get_analysis_modules()
modules = detection_modules + analysis_modules
if options.list_plugins:
list_plugins(modules)
return 0
options.all_enabled_modules = [
module for module in modules if module.is_enabled(options)
]
# converts from a namespace to an antismash.config.Config instance so
# modules can't fiddle with it
options = update_config(options)
if options.check_prereqs_only:
try:
check_prerequisites(modules)
except RuntimeError:
print("Some module prerequisites not satisfied")
return 1
print("All prerequisites satisfied")
return 0
else:
check_prerequisites(options.all_enabled_modules)
# start up profiling if relevant
if options.profile:
profiler = cProfile.Profile()
profiler.enable()
# ensure the provided options are valid
if not verify_options(options, options.all_enabled_modules):
return 1 # TODO: change to a raise?
# check that at least one module will run
if not options.all_enabled_modules:
raise ValueError("No detection or analysis modules enabled")
start_time = datetime.now()
results = read_data(sequence_file, options)
# reset module timings
results.timings_by_record = {}
prepare_output_directory(options.output_dir, sequence_file
or options.reuse_results)
results.records = record_processing.pre_process_sequences(
results.records, options, genefinding)
for seq_record, previous_result in zip(results.records, results.results):
# skip if we're not interested in it
if seq_record.skip:
continue
timings = run_detection(seq_record, options, previous_result)
# and skip analysis if detection didn't find anything
if not seq_record.get_clusters():
continue
analysis_timings = analyse_record(seq_record, options,
analysis_modules, previous_result)
timings.update(analysis_timings)
results.timings_by_record[seq_record.id] = timings
# Write results
json_filename = os.path.join(options.output_dir, results.input_file)
json_filename = os.path.splitext(json_filename)[0] + ".json"
logging.debug("Writing json results to '%s'", json_filename)
results.write_to_file(json_filename)
# now that the json is out of the way, annotate the record
# otherwise we could double annotate some areas
annotate_records(results)
# create relevant output files
write_outputs(results, options)
# save profiling data
if options.profile:
profiler.disable()
write_profiling_results(
profiler, os.path.join(options.output_dir, "profiling_results"))
running_time = datetime.now() - start_time
# display module runtimes before total time
if options.debug:
log_module_runtimes(results.timings_by_record)
logging.debug("antiSMASH calculation finished at %s; runtime: %s",
str(datetime.now()), str(running_time))
logging.info("antiSMASH status: SUCCESS")
return 0
def run_on_records(self, records):
record_processing.pre_process_sequences(records, self.options,
self.genefinding)