def test_min_genes_required_cfg(self):
# test min_genes_required is specified from configuration
# so this value must overload the value read from xml
def_2_parse = set()
model_fqn = 'foo/model_5'
def_2_parse.add(model_fqn)
parsed = set()
min_genes_required = [[model_fqn, '4']]
self.args.min_genes_required = min_genes_required
self.cfg = Config(MacsyDefaults(), self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.min_genes_required, 4)
def test_get_def_to_detect(self):
cmd_args = argparse.Namespace()
cmd_args.models_dir = os.path.join(self._data_dir, 'fake_model_dir')
cmd_args.models = ('set_1', 'def_1_1', 'def_1_2', 'def_1_3')
registry = ModelRegistry()
models_location = scan_models_dir(cmd_args.models_dir)
for ml in models_location:
registry.add(ml)
# case where models are specified on command line
res, model_family, model_vers = get_def_to_detect(
('set_1', ['def_1_1', 'def_1_2', 'def_1_3']), registry)
model_loc = registry['set_1']
self.assertEqual(model_family, 'set_1')
self.assertEqual(model_vers, '0.0b2')
exp = [
model_loc.get_definition(name)
for name in ('set_1/def_1_1', 'set_1/def_1_2', 'set_1/def_1_3')
]
self.assertListEqual(res, exp)
# case we search all models
res, model_family, model_vers = get_def_to_detect(('set_1', ['all']),
registry)
self.assertEqual(model_family, 'set_1')
self.assertEqual(model_vers, '0.0b2')
exp = model_loc.get_all_definitions()
self.assertListEqual(res, exp)
# case the models required does not exists
with self.assertRaises(ValueError):
get_def_to_detect(('set_1', ['FOO', 'BAR']), registry)
def test_inter_gene_max_space_cfg(self):
# test inter_gene_max_space is specified from configuration
# so this value must overload the value read from xml
model_fqn = 'foo/model_5'
inter_gene_max_space_cfg = [[model_fqn, '222']]
self.args.inter_gene_max_space = inter_gene_max_space_cfg
self.cfg = Config(MacsyDefaults(), self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.inter_gene_max_space, 222)
def test_add_get(self):
mr = ModelRegistry()
model_complex_expected = ModelLocation(path=self.complex_dir)
with self.assertRaises(KeyError) as ctx:
mr[model_complex_expected.name]
self.assertEqual(str(ctx.exception),
'"No such model definition: \'complex\'"')
mr.add(model_complex_expected)
self.assertEqual(model_complex_expected,
mr[model_complex_expected.name])
def list_models(args):
"""
:param args: The command line argument once parsed
:type args: :class:`argparse.Namespace` object
:return: a string representation of all models and submodels installed.
:rtype: str
"""
config = Config(MacsyDefaults(), args)
registry = ModelRegistry()
models_loc_available = scan_models_dir(
config.models_dir(),
profile_suffix=config.profile_suffix(),
relative_path=config.relative_path())
for model_loc in models_loc_available:
registry.add(model_loc)
return str(registry)
def _find_all_installed_packages() -> ModelRegistry:
"""
:return: all models installed
"""
defaults = MacsyDefaults()
config = Config(defaults, argparse.Namespace())
system_model_dir = config.models_dir()
user_model_dir = os.path.join(os.path.expanduser('~'), '.macsyfinder', 'data')
model_dirs = (system_model_dir, user_model_dir) if os.path.exists(user_model_dir) else (system_model_dir,)
registry = ModelRegistry()
for model_dir in model_dirs:
try:
for model_loc in scan_models_dir(model_dir, profile_suffix=config.profile_suffix):
registry.add(model_loc)
except PermissionError as err:
_log.warning(f"{model_dir} is not readable: {err} : skip it.")
return registry
def setUp(self):
defaults = MacsyDefaults()
self.args = argparse.Namespace()
self.args.sequence_db = self.find_data("base", "test_1.fasta")
self.args.db_type = 'gembase'
self.args.models_dir = self.find_data('models')
self.args.res_search_dir = tempfile.gettempdir()
self.cfg = Config(defaults, self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.profile_factory = ProfileFactory(self.cfg)
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
def _find_all_installed_packages(models_dir=None) -> ModelRegistry:
"""
:return: all models installed
"""
defaults = MacsyDefaults()
args = argparse.Namespace()
if models_dir is not None:
args.models_dir = models_dir
config = Config(defaults, args)
model_dirs = config.models_dir()
registry = ModelRegistry()
for model_dir in model_dirs:
try:
for model_loc in scan_models_dir(
model_dir, profile_suffix=config.profile_suffix()):
registry.add(model_loc)
except PermissionError as err:
_log.warning(f"{model_dir} is not readable: {err} : skip it.")
return registry
def test_models(self):
mr = ModelRegistry()
model_complex_expected = ModelLocation(path=self.complex_dir)
model_simple_expected = ModelLocation(path=self.simple_dir)
mr.add(model_complex_expected)
mr.add(model_simple_expected)
models_received = sorted(mr.models())
models_expected = sorted(
[model_complex_expected, model_simple_expected])
self.assertListEqual(models_received, models_expected)
def test_max_nb_genes_cfg(self):
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
# max_nb_genes is specified in xml
# no user configuration on this
self.cfg = Config(MacsyDefaults(), self.args)
model_fqn = 'foo/model_6' # 4 genes in this model but xml specify 3
self.cfg = Config(MacsyDefaults(), self.args)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.max_nb_genes, 3)
# max_nb_genes is specified from configuration
# so this value must overload the value read from xml
model_fqn = 'foo/model_5' # 4 genes in this model
max_nb_genes = [[model_fqn, '6']]
self.args.max_nb_genes = max_nb_genes
self.cfg = Config(MacsyDefaults(), self.args)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.max_nb_genes, 6)
def test_multi_loci_cfg(self):
# test multi_loci is specified from configuration
# so this value must overload the value read from xml
model_fqn = 'foo/model_5'
self.args.multi_loci = model_fqn
self.cfg = Config(MacsyDefaults(), self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertTrue(m.multi_loci)
def test_str(self):
mr = ModelRegistry()
model_complex_expected = ModelLocation(path=self.complex_dir)
model_simple_expected = ModelLocation(path=self.simple_dir)
mr.add(model_complex_expected)
mr.add(model_simple_expected)
expected_output = """complex
/subdef_1
/def_1_1
/def_1_2
/subdef_2
/def_2_1
/def_2_2
simple
/def_1
/def_2
"""
self.assertEqual(expected_output, str(mr))
class TestModelParser(MacsyTest):
def setUp(self):
defaults = MacsyDefaults()
self.args = argparse.Namespace()
self.args.sequence_db = self.find_data("base", "test_1.fasta")
self.args.db_type = 'gembase'
self.args.models_dir = self.find_data('models')
self.args.res_search_dir = tempfile.gettempdir()
self.cfg = Config(defaults, self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.profile_factory = ProfileFactory(self.cfg)
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
def tearDown(self):
try:
shutil.rmtree(self.cfg.working_dir())
except:
pass
def test_parse_with_exchangeable(self):
model_name = 'model_1'
model_family = 'foo'
fqn = f"{model_family}/{model_name}"
#def_2_parse = set()
#def_2_parse.add(fqn)
models_2_detect = [self.model_registry['foo'].get_definition(fqn)]
self.parser.parse(models_2_detect)
self.assertEqual(len(self.model_bank), 1)
m1 = self.model_bank[fqn]
self.assertEqual(m1.name, model_name)
self.assertEqual(m1.fqn, fqn)
self.assertEqual(m1.inter_gene_max_space, 20)
self.assertEqual(m1.min_mandatory_genes_required, 2)
self.assertEqual(m1.min_genes_required, 4)
self.assertTrue(m1.multi_loci)
self.assertEqual(len(m1.mandatory_genes), 2)
mandatory_genes_name = sorted([g.name for g in m1.mandatory_genes])
theoric_list = sorted(["sctJ_FLG", "sctN_FLG"])
self.assertListEqual(mandatory_genes_name, theoric_list)
self.assertEqual(len(m1.accessory_genes), 2)
accessory_genes_name = sorted([g.name for g in m1.accessory_genes])
theoric_list = sorted(["flgB", "flgC"])
self.assertListEqual(accessory_genes_name, theoric_list)
self.assertEqual(len(m1.neutral_genes), 2)
neutral_genes_name = sorted([g.name for g in m1.neutral_genes])
theoric_list = sorted(["fliE", "tadZ"])
self.assertListEqual(neutral_genes_name, theoric_list)
self.assertEqual(len(m1.forbidden_genes), 1)
forbidden_genes_name = sorted([g.name for g in m1.forbidden_genes])
theoric_list = sorted(["sctC"])
self.assertListEqual(forbidden_genes_name, theoric_list)
sctJ_FLG = m1.get_gene('sctJ_FLG')
sctJ_FLG_exchangeables = sctJ_FLG.exchangeables
self.assertEqual(len(sctJ_FLG_exchangeables), 2)
self.assertEqual(sctJ_FLG_exchangeables[0].name, 'sctJ')
self.assertEqual(sctJ_FLG_exchangeables[1].name, 'abc')
self.assertTrue(isinstance(sctJ_FLG_exchangeables[0], Exchangeable))
self.assertTrue(isinstance(sctJ_FLG_exchangeables[0]._gene, CoreGene))
self.assertTrue(
isinstance(sctJ_FLG_exchangeables[0].alternate_of(), ModelGene))
self.assertTrue(sctJ_FLG_exchangeables[0].loner)
self.assertFalse(sctJ_FLG.is_exchangeable)
sctJ = m1.get_gene('sctJ')
self.assertTrue(sctJ.is_exchangeable)
def test_exchangeable_inheritance(self):
def_2_parse = set()
def_2_parse.add('foo/model_1')
models_2_detect = [
self.model_registry['foo'].get_definition('foo/model_1')
]
self.parser.parse(models_2_detect)
m1 = self.model_bank['foo/model_1']
sctJ = m1.get_gene('sctJ')
self.assertTrue(sctJ.is_exchangeable)
self.assertTrue(sctJ.loner)
self.assertTrue(sctJ.multi_system)
self.assertFalse(sctJ.multi_model)
sctJ_FLG = m1.get_gene('sctJ_FLG')
self.assertTrue(sctJ_FLG.multi_system)
abc = m1.get_gene('abc')
self.assertFalse(abc.multi_system)
sctN = m1.get_gene('sctN')
sctN_FLG = m1.get_gene('sctN_FLG')
self.assertFalse(sctN_FLG.loner)
self.assertTrue(sctN.loner)
self.assertIsNone(sctN_FLG.inter_gene_max_space)
self.assertEqual(sctN.inter_gene_max_space, 10)
self.assertFalse(sctN_FLG.multi_model)
self.assertFalse(sctN.multi_model)
gspD = m1.get_gene('gspD')
self.assertFalse(sctN_FLG.multi_system)
self.assertTrue(gspD.multi_model)
self.assertTrue(gspD.multi_system)
def test_model_w_unkown_attr(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/model_w_unknown_attribute')
]
with self.assertRaises(MacsypyError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"unable to parse model definition 'foo/model_w_unknown_attribute' : "
"The model definition model_w_unknown_attribute.xml has an unknow attribute 'multi-loci'. "
"Please fix the definition.")
def test_gene_w_unkown_attr(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/gene_w_unknown_attribute')
]
with self.assertRaises(MacsypyError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"unable to parse model definition 'foo/gene_w_unknown_attribute' : "
"The model definition gene_w_unknown_attribute.xml has an unknown attribute 'multi-system' for a gene."
" Please fix the definition.")
def test_wo_presence(self):
model_2_detect = [
self.model_registry['foo'].get_definition('foo/fail_wo_presence')
]
with self.assertRaises(SyntaxError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"Invalid model definition 'foo/fail_wo_presence': gene 'sctN_FLG' without presence"
)
def test_invalid_presence(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/fail_invalid_presence')
]
with self.assertRaises(SyntaxError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"Invalid model 'fail_invalid_presence' definition: presence value must be either: "
"'mandatory', 'accessory', 'neutral', 'forbidden' not foo_bar")
def test_gene_no_name(self):
model_2_detect = [
self.model_registry['foo'].get_definition('foo/gene_no_name')
]
with self.assertRaises(SyntaxError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"Invalid model definition 'foo/gene_no_name': gene without name")
def test_invalid_homolog(self):
model_2_detect = [
self.model_registry['foo'].get_definition('foo/invalid_homolog')
]
with self.assertRaises(MacsypyError) as context:
self.parser.parse(model_2_detect)
self.assertEqual(str(context.exception),
"'foo/foo_bar': No such profile")
def test_invalid_homolog_2(self):
model_2_detect = [
self.model_registry['foo'].get_definition('foo/invalid_homolog_2')
]
with self.assertRaises(SyntaxError) as ctx:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(ctx.exception),
"Invalid model definition 'foo/invalid_homolog_2': gene without name"
)
def test_bad_min_genes_required(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/bad_min_genes_required')
]
with self.assertRaises(ModelInconsistencyError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
'model \'bad_min_genes_required\' is not consistent: min_genes_required 16 must be lesser '
'or equal than the number of "accessory" and "mandatory" components in the model: 6'
)
def test_bad_min_genes_required_2(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/bad_min_genes_required_2')
]
with self.catch_log():
with self.assertRaisesRegex(
SyntaxError, "Invalid model definition (.*): "
"min_genes_required must be an integer: 16.5"):
self.parser.parse(model_2_detect)
def test_bad_min_mandatory_genes_required(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/bad_min_mandatory_genes_required')
]
with self.catch_log():
with self.assertRaises(ModelInconsistencyError) as context:
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
'model \'bad_min_mandatory_genes_required\' is not consistent: min_genes_required 16 must '
'be lesser or equal than the number of "accessory" and "mandatory" components in the model: 6'
)
def test_bad_min_mandatory_genes_required_2(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/bad_min_mandatory_genes_required_2')
]
with self.assertRaises(ModelInconsistencyError) as context:
with self.catch_log():
# error raised by System initialization
# which occur before check_consistency
# the last test : not(model.min_mandatory_genes_required <= model.min_genes_required)
# seems to be useless
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"foo/bad_min_mandatory_genes_required_2: min_genes_required '6' must be greater or equal"
" than min_mandatory_genes_required '8'")
def test_bad_min_mandatory_genes_required_4(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/bad_min_mandatory_genes_required_4')
]
with self.assertRaisesRegex(
SyntaxError, "Invalid model definition (.*): "
"min_mandatory_genes_required must be an integer: 12.5"):
with self.catch_log():
self.parser.parse(model_2_detect)
def test_min_mandatory_genes_required_lesser_than_mandatory_genes(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/bad_min_mandatory_genes_required_3')
]
with self.assertRaises(ModelInconsistencyError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"model 'bad_min_mandatory_genes_required_3' is not consistent:"
" 'min_mandatory_genes_required': 6 must be lesser or equal than the number of 'mandatory' "
"components in the model: 5")
def test_only_one_accessory(self):
model_2_detect = [
self.model_registry['foo'].get_definition('foo/only_one_accessory')
]
with self.assertRaises(ModelInconsistencyError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(str(context.exception),
f"model 'only_one_accessory' is not consistent: there is only one gene in your model. " \
f"So its status should be 'mandatory'.")
def test_bad_max_nb_genes(self):
model_2_detect = [
self.model_registry['foo'].get_definition('foo/bad_max_nb_genes')
]
with self.assertRaises(SyntaxError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
model_name, def_name = model_2_detect[0].split_fqn(
model_2_detect[0].fqn)
self.assertEqual(
str(context.exception),
"Invalid model definition ({0}.xml): max_nb_genes must be an integer: HOHOHO"
.format(
os.path.join(self.cfg.models_dir()[0], model_name,
'definitions', def_name)))
def test_bad_inter_gene_max_space(self):
fqn = 'foo/bad_inter_gene_max_space'
model_family, model_name = fqn.split('/')
model_2_detect = [self.model_registry['foo'].get_definition(fqn)]
with self.assertRaises(SyntaxError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"Invalid model definition ({}): inter_gene_max_space must be an integer: 12.5"
.format(
os.path.join(self.cfg.models_dir()[0], model_family,
'definitions', model_name + ".xml")))
def test_no_inter_gene_max_space(self):
model_2_detect = [
self.model_registry['foo'].get_definition(
'foo/no_inter_gene_max_space')
]
with self.assertRaises(SyntaxError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"Invalid model definition ({}): inter_gene_max_space must be defined"
.format(
os.path.join(self.cfg.models_dir()[0],
"foo/definitions/no_inter_gene_max_space.xml")))
def test_loner(self):
model_fqn = 'foo/model_5'
model_2_detect = [self.model_registry['foo'].get_definition(model_fqn)]
self.parser.parse(model_2_detect)
m5 = self.model_bank[model_fqn]
m5_flgC = m5.get_gene('flgC')
self.assertFalse(m5_flgC.loner)
m5_tadZ = m5.get_gene('tadZ')
self.assertTrue(m5_tadZ.loner)
model_fqn = 'foo/model_6'
model_2_detect = [self.model_registry['foo'].get_definition(model_fqn)]
self.parser.parse(model_2_detect)
m6 = self.model_bank[model_fqn]
m6_flgC = m6.get_gene('flgC')
self.assertFalse(m6_flgC.loner)
m6_tadZ = m6.get_gene('tadZ')
self.assertFalse(m6_tadZ.loner)
def test_multi_system(self):
model_fqn = 'foo/model_5'
model_2_detect = [self.model_registry['foo'].get_definition(model_fqn)]
self.parser.parse(model_2_detect)
m = self.model_bank[model_fqn]
flgC = m.get_gene('flgC')
self.assertFalse(flgC.multi_system)
fliE = m.get_gene('fliE')
self.assertTrue(fliE.multi_system)
def test_multi_model(self):
model_fqn = 'foo/model_5'
model_2_detect = [self.model_registry['foo'].get_definition(model_fqn)]
self.parser.parse(model_2_detect)
m = self.model_bank[model_fqn]
flgC = m.get_gene('flgC')
self.assertFalse(flgC.multi_model)
abc = m.get_gene('abc')
self.assertTrue(abc.multi_model)
def test_gene_inter_gene_max_space(self):
model_fqn = ['foo/model_5', 'foo/model_6']
models_2_detect = [
self.model_registry['foo'].get_definition(fqn) for fqn in model_fqn
]
self.parser.parse(models_2_detect)
m5 = self.model_bank['foo/model_5']
self.assertEqual(m5.name, 'model_5')
self.assertEqual(m5.fqn, 'foo/model_5')
self.assertEqual(m5.inter_gene_max_space, 20)
m5_flgB = m5.get_gene('flgB')
m5_flgC = m5.get_gene('flgC')
self.assertIsNone(m5_flgB.inter_gene_max_space)
self.assertEqual(m5_flgC.inter_gene_max_space, 2)
m6 = self.model_bank['foo/model_6']
m6_flgC = m6.get_gene('flgC')
self.assertEqual(m6_flgC.inter_gene_max_space, 12)
def test_inter_gene_max_space_cfg(self):
# test inter_gene_max_space is specified from configuration
# so this value must overload the value read from xml
model_fqn = 'foo/model_5'
inter_gene_max_space_cfg = [[model_fqn, '222']]
self.args.inter_gene_max_space = inter_gene_max_space_cfg
self.cfg = Config(MacsyDefaults(), self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.inter_gene_max_space, 222)
def test_min_mandatory_genes_required_cfg(self):
# test min_mandatory_genes_required is specified from configuration
# so this value must overload the value read from xml
model_fqn = 'foo/model_5'
min_mandatory_genes_required = [[model_fqn, '3']]
self.args.min_mandatory_genes_required = min_mandatory_genes_required
self.cfg = Config(MacsyDefaults(), self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.min_mandatory_genes_required, 3)
def test_min_genes_required_cfg(self):
# test min_genes_required is specified from configuration
# so this value must overload the value read from xml
def_2_parse = set()
model_fqn = 'foo/model_5'
def_2_parse.add(model_fqn)
parsed = set()
min_genes_required = [[model_fqn, '4']]
self.args.min_genes_required = min_genes_required
self.cfg = Config(MacsyDefaults(), self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.min_genes_required, 4)
def test_max_nb_genes_cfg(self):
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
# max_nb_genes is specified in xml
# no user configuration on this
self.cfg = Config(MacsyDefaults(), self.args)
model_fqn = 'foo/model_6' # 4 genes in this model but xml specify 3
self.cfg = Config(MacsyDefaults(), self.args)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.max_nb_genes, 3)
# max_nb_genes is specified from configuration
# so this value must overload the value read from xml
model_fqn = 'foo/model_5' # 4 genes in this model
max_nb_genes = [[model_fqn, '6']]
self.args.max_nb_genes = max_nb_genes
self.cfg = Config(MacsyDefaults(), self.args)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertEqual(m.max_nb_genes, 6)
def test_multi_loci_cfg(self):
# test multi_loci is specified from configuration
# so this value must overload the value read from xml
model_fqn = 'foo/model_5'
self.args.multi_loci = model_fqn
self.cfg = Config(MacsyDefaults(), self.args)
self.model_bank = ModelBank()
self.gene_bank = GeneBank()
self.model_registry = ModelRegistry()
models_location = scan_models_dir(self.args.models_dir)
for ml in models_location:
self.model_registry.add(ml)
self.parser = DefinitionParser(self.cfg, self.model_bank,
self.gene_bank, self.model_registry,
self.profile_factory)
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
self.parser.parse(models_2_detect)
m = self.model_bank[model_fqn]
self.assertTrue(m.multi_loci)
def test_bad_gene_inter_gene_max_space_2(self):
model_fqn = 'foo/bad_inter_gene_max_space_2'
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
with self.assertRaises(SyntaxError) as ctx:
with self.catch_log():
self.parser.parse(models_2_detect)
self.assertEqual(
str(ctx.exception),
"Invalid model definition 'bad_inter_gene_max_space_2': "
"inter_gene_max_space must be an integer: 2.5")
def test_bad_exchangeable_inter_gene_max_space(self):
fqn = 'foo/bad_exchangeable_inter_gene_max_space'
model_2_detect = [self.model_registry['foo'].get_definition(fqn)]
with self.assertRaises(SyntaxError) as context:
with self.catch_log():
self.parser.parse(model_2_detect)
self.assertEqual(
str(context.exception),
"Invalid model definition 'bad_exchangeable_inter_gene_max_space': "
"inter_gene_max_space must be an integer: 1.5")
def test_parse_model_old_syntax(self):
# the attribute vers is not set
model_fqn = 'foo/model_old_1'
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
with self.catch_log(log_name='macsypy') as log:
with self.assertRaises(MacsypyError) as ctx:
self.parser.parse(models_2_detect)
log_msg = log.get_value().strip()
self.assertEqual(
log_msg, "unable to parse model definition 'foo/model_old_1' : "
"The model definition model_old_1.xml is not versioned. Please update your model."
)
# the root is system instead of mmodel
model_fqn = 'foo/model_old_2'
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
with self.catch_log(log_name='macsypy') as log:
with self.assertRaises(MacsypyError) as ctx:
self.parser.parse(models_2_detect)
log_msg = log.get_value().strip()
self.assertEqual(
log_msg, f"unable to parse model definition '{model_fqn}' : "
"The model definition model_old_2.xml is obsolete. Please update your model."
)
# there still system_ref attribute
model_fqn = 'foo/model_old_3'
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
with self.catch_log(log_name='macsypy') as log:
with self.assertRaises(MacsypyError) as ctx:
self.parser.parse(models_2_detect)
log_msg = log.get_value().strip()
self.assertEqual(
log_msg, f"unable to parse model definition '{model_fqn}' : "
"The model definition model_old_3.xml is obsolete. Please update your model."
)
# there still homologs tag
model_fqn = 'foo/model_old_4'
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
with self.catch_log(log_name='macsypy') as log:
with self.assertRaises(MacsypyError) as ctx:
self.parser.parse(models_2_detect)
log_msg = log.get_value().strip()
self.assertEqual(
log_msg, f"unable to parse model definition '{model_fqn}' : "
"The model definition model_old_4.xml is obsolete. Please update your model."
)
# there still analogs tag
model_fqn = 'foo/model_old_5'
models_2_detect = [
self.model_registry['foo'].get_definition(model_fqn)
]
with self.catch_log(log_name='macsypy') as log:
with self.assertRaises(MacsypyError) as ctx:
self.parser.parse(models_2_detect)
log_msg = log.get_value().strip()
self.assertEqual(
log_msg, f"unable to parse model definition '{model_fqn}' : "
"The model definition model_old_5.xml is obsolete. Please update your model."
)
def search_systems(config, model_bank, gene_bank, profile_factory, logger):
"""
Do the job, this function is the orchestrator of all the macsyfinder mechanics
at the end several files are produced containing the results
- macsyfinder.conf: The set of variables used to runt this job
- macsyfinder.systems: The list of the potential systems
- macsyfinder.rejected_cluster: The list of all clusters and clustrs combination
which has been rejected and the reason
- macsyfinder.log: the copy of the standard output
:param config: The MacSyFinder Configuration
:type config: :class:`macsypy.config.Config` object
:param model_bank: The bank populated with the available models
:type model_bank: :class:`macsypy.model.ModelBank` object
:param gene_bank: the bank containing all genes
:type gene_bank: :class:`macsypy.gene.GeneBank` object
:param profile_factory: The profile factory
:type profile_factory: :class:`macsypy.gene.ProfileFactory`
:param logger: The logger use to display information to the user.
It must be initialized. see :func:`macsypy.init_logger`
:type logger: :class:`colorlog.Logger` object
:return: the systems and rejected clusters found
:rtype: ([:class:`macsypy.system.System`, ...], [:class:`macsypy.cluster.RejectedCluster`, ...])
"""
working_dir = config.working_dir()
config.save(path_or_buf=os.path.join(working_dir, config.cfg_name))
registry = ModelRegistry()
models_loc_available = scan_models_dir(
config.models_dir(),
profile_suffix=config.profile_suffix(),
relative_path=config.relative_path())
for model_loc in models_loc_available:
registry.add(model_loc)
# build indexes
idx = Indexes(config)
idx.build(force=config.idx)
# create models
parser = DefinitionParser(config, model_bank, gene_bank, registry,
profile_factory)
try:
models_def_to_detect = get_def_to_detect(config.models(), registry)
except KeyError as err:
sys.exit(f"macsyfinder: {err}")
parser.parse(models_def_to_detect)
logger.info(
f"MacSyFinder's results will be stored in working_dir{working_dir}")
logger.info(f"Analysis launched on {config.sequence_db()} for model(s):")
for m in models_def_to_detect:
logger.info(f"\t- {m.fqn}")
models_to_detect = [
model_bank[model_loc.fqn] for model_loc in models_def_to_detect
]
all_genes = []
for model in models_to_detect:
genes = model.mandatory_genes + model.accessory_genes + model.neutral_genes + model.forbidden_genes
# Exchangeable (formerly homologs/analogs) are also added because they can "replace" an important gene...
ex_genes = []
for g in genes:
ex_genes += g.exchangeables
all_genes += (genes + ex_genes)
#############################################
# this part of code is executed in parallel
#############################################
try:
all_reports = search_genes(all_genes, config)
except Exception as err:
sys.exit(str(err))
#############################################
# end of parallel code
#############################################
all_hits = [
hit for subl in [report.hits for report in all_reports] for hit in subl
]
if len(all_hits) > 0:
# It's important to keep this sorting to have in last all_hits version
# the hits with the same replicon_name and position sorted by score
# the best score in first
hits_by_replicon = {}
for hit in all_hits:
if hit.replicon_name in hits_by_replicon:
hits_by_replicon[hit.replicon_name].append(hit)
else:
hits_by_replicon[hit.replicon_name] = [hit]
for rep_name in hits_by_replicon:
hits_by_replicon[rep_name] = get_best_hits(
hits_by_replicon[rep_name], key='score')
hits_by_replicon[rep_name].sort(key=attrgetter('position'))
models_to_detect = sorted(models_to_detect, key=attrgetter('name'))
db_type = config.db_type()
if db_type in ('ordered_replicon', 'gembase'):
systems, rejected_clusters = _search_in_ordered_replicon(
hits_by_replicon, models_to_detect, config, logger)
return systems, rejected_clusters
elif db_type == "unordered":
likely_systems, rejected_hits = _search_in_unordered_replicon(
hits_by_replicon, models_to_detect, logger)
return likely_systems, rejected_hits
else:
assert False, f"dbtype have an invalid value {db_type}"
else:
# No hits detected
return [], []