def test_generate_sbml_model():
# Test a non-reversible reaction
s1 = Species("S1")
s2 = Species("S2")
rx0 = Reaction.from_massaction(inputs=[s1], outputs=[s2], k_forward=0.1)
crn = ChemicalReactionNetwork(species=[s1, s2], reactions=[rx0])
# generate an sbml model
document, model = crn.generate_sbml_model()
# all species from the CRN are accounted for
assert len(model.getListOfSpecies()) == len(crn.species)
# all reactions from the CRN are accounted for
assert len(model.getListOfReactions()) == len(crn.reactions)
document2, model2 = crn.generate_sbml_model(check_validity=False)
assert len(model2.getListOfSpecies()) == len(
crn.species) #the same model is made if you dont check validity
assert len(model2.getListOfReactions()) == len(
crn.reactions) #the same model is made if you dont check validity
# reversible needs to be off!
# assert not model.getListOfReactions()[0].isSetReversible()
# test a reversible reaction
rx1 = Reaction.from_massaction(inputs=[s1],
outputs=[s2],
k_forward=0.1,
k_reverse=0.1)
rxn_list = [rx1]
crn = ChemicalReactionNetwork(species=[s1, s2], reactions=rxn_list)
# generate an sbml model
document, model = crn.generate_sbml_model()
# all species from the CRN are accounted for
assert len(model.getListOfSpecies()) == len(crn.species)
# all reactions from the CRN are accounted for
assert len(model.getListOfReactions()) == 2
# although sbml represents a reverisble reaction with reversible flag
# BioCRNpyler always creates two reactions, because this is correct
# for stochastic simulation with SBML.
sbml_rxn = model.getListOfReactions()
# assert not sbml_rxn[0].isSetReversible()
# assert not sbml_rxn[1].isSetReversible()
# Test propagation for for_bioscrape keyword
# generate an sbml model with for_bioscrape = True
document, model = crn.generate_sbml_model(for_bioscrape=True)
for r in model.getListOfReactions():
assert r.getAnnotation() is not None
assert validate_sbml(document) == 0
# generate an sbml model with for_bioscrape = False
document, model = crn.generate_sbml_model(for_bioscrape=False)
for r in model.getListOfReactions():
assert r.getAnnotation() is None
assert validate_sbml(document) == 0
def test_generate_sbml_model_parameter_names():
s1 = Species("S1")
s2 = Species("S2")
# The correct parameter formating should be: "name_partid_mechanism"
key0 = ParameterKey(mechanism="m", part_id="p", name="n")
k0 = ParameterEntry("n", 1.0, parameter_key=key0)
# Tests 3 Parameter entries with seemingly redundant keys.
# In SBML parameter id, None will be kept blank, resulting in
# different numbers of underscores between the V's
k1 = ParameterEntry("v", 1.0, parameter_key=None)
key2 = ParameterKey(mechanism=None, part_id="v", name="v")
k2 = ParameterEntry("v", 2.0, parameter_key=key2)
key3 = ParameterKey(mechanism="v", part_id=None, name="v")
k3 = ParameterEntry("v", 2.0, parameter_key=key3)
key4 = ParameterKey(mechanism="v", part_id="v", name="v")
k4 = ParameterEntry("v", 2.0, parameter_key=key4)
# Throw a duplicate key for good measure!
rx0 = Reaction.from_massaction(inputs=[],
outputs=[s1],
k_forward=k0,
k_reverse=k1)
rx1 = Reaction.from_massaction(inputs=[s1],
outputs=[s2],
k_forward=k1,
k_reverse=k2)
rx2 = Reaction.from_massaction(inputs=2 * [s1],
outputs=2 * [s2],
k_forward=k3,
k_reverse=k4)
rxn_list = [rx0, rx1, rx2]
crn = ChemicalReactionNetwork(species=[s1, s2], reactions=rxn_list)
document, model = crn.generate_sbml_model()
assert validate_sbml(document) == 0
correct_ids = set(["v_v_", "v__v", "v_v_v", "v__", "n_p_m"])
ids = set([p.getId() for p in model.getListOfParameters()])
assert ids == correct_ids
propensity_types = [
'hillpositive', 'proportionalhillpositive', 'hillnegative',
'proportionalhillnegative', 'massaction', 'general'
]
kb = 100
ku = 10
kex = 1.
kd = .1
G = Species(name="G", material_type="dna") #DNA
A = Species(name="A", material_type="protein") #Activator
GA = ComplexSpecies([G, A, A]) #Activated Gene
X = Species(name="X", material_type="protein")
rxnd = Reaction([X], [], kd)
#Massaction Unregulated
species1 = [G, A, GA, X]
rxn0_1 = Reaction([G, A, A], [GA], k=kb, k_rev=ku)
rxn0_2 = Reaction([GA], [GA, X], k=kex)
CRN0 = ChemicalReactionNetwork(species1, [rxn0_1, rxn0_2, rxnd])
rxn1_1 = Reaction([G, A, A], [GA], k=kb, k_rev=ku)
rxn1_2 = Reaction([G], [G, X], k=kex)
CRN1 = ChemicalReactionNetwork(species1, [rxn1_1, rxn1_2, rxnd])
#hill positive
species2 = [G, A, X]
rxn2_1 = Reaction([G], [G, X],
def test_add_reaction():
"""
Test adding reaction to SBML for combinatorially created list of models
"""
#create species
S1, S2, S3 = Species("S1"), Species("S2"), Species("S3")
species = [S1, S2, S3]
#create some parameters
key1 = ParameterKey(name="k", mechanism="m", part_id="pid")
k1 = ParameterEntry("k1", 1.11, key1)
k2 = ParameterEntry("k2", 2.22)
stochiometries = [([], [S1]), ([S1], []), ([S1], [S2]),
(2 * [S1], [S2, S3])]
propensities = [
MassAction(k_forward=1.),
MassAction(k_forward=1, k_reverse=.1),
MassAction(k_forward=k1),
MassAction(k_forward=k1, k_reverse=k2),
HillPositive(k=1, n=2., K=3., s1=S1),
HillPositive(k=k1, n=2., K=k2, s1=S1),
HillNegative(k=1, n=2., K=3., s1=S1),
HillNegative(k=k1, n=k2, K=3., s1=S1),
ProportionalHillPositive(k=1, n=2, K=3., s1=S1, d=S2),
ProportionalHillPositive(k=k1, n=2, K=k2, s1=S1, d=S2),
ProportionalHillNegative(k=1, n=2, K=3., s1=S1, d=S2),
ProportionalHillNegative(k=k1, n=2, K=k2, s1=S1, d=S2),
GeneralPropensity('k1*2 - k2/S1^2',
propensity_species=[S1],
propensity_parameters=[k1, k2]),
GeneralPropensity('S1^2 + S2^2 + S3^2',
propensity_species=[S1, S2, S3],
propensity_parameters=[])
]
for prop in propensities: #Cycle through different propensity types
for inputs, outputs in stochiometries: #cycle through different stochiometries
for stochastic in [True, False]: #Toggle Stochastic
rxn_num = 0
model_id = f"{prop.name}_model_with_stochastic_{stochastic}"
document, model = create_sbml_model(model_id=model_id)
add_all_species(model, species)
rxn = Reaction(inputs, outputs,
propensity_type=prop) #create a reaction
try:
sbml_reaction = add_reaction(
model, rxn, f"r{rxn_num}",
stochastic=stochastic) #add reaction to SBML Model
rxn_num += 1 #increment reaction id
crn_reactants = [str(s) for s in inputs]
crn_products = [str(s) for s in outputs]
sbml_products = [
p.getSpecies()
for p in sbml_reaction.getListOfProducts()
]
sbml_reactants = [
p.getSpecies()
for p in sbml_reaction.getListOfReactants()
]
#test that the reaction has the right inputs and outputs
assert set(crn_reactants) == set(sbml_reactants)
assert set(crn_products) == set(sbml_products)
if len(crn_reactants) > 0:
assert all(
crn_reactants.count(s.getSpecies()) == int(
s.getStoichiometry())
for s in sbml_reaction.getListOfReactants())
if len(crn_products) > 0:
assert all(
crn_products.count(s.getSpecies()) == int(
s.getStoichiometry())
for s in sbml_reaction.getListOfProducts())
assert not sbml_reaction.getReversible()
#TODO is there a smart way to test that the rate formula is correct?
assert validate_sbml(
document) == 0 # Validate the SBML model
except Exception as e: #collect errors to display
error_txt = f"Unexpected Error: in sbmlutil.add_reaction {rxn} for {model_id}. \n {str(e)}."
raise Exception(error_txt)
def test_sbml_basics():
"""
This test aims to test all Python libSBML functions and their return values.
It will catch errors if certain libSBML functions are not working as expected.
For example: At times, if the reaction rate law string is mis-formatted, it is
ignored by the SBML writing code and not caught by any other tests. Similarly, this
test will catch modifier reference, parameter values etc.
"""
def check(value, message):
"""If 'value' is None, prints an error message constructed using
'message' and then exits with status code 1 (for libsbml). If 'value' is an integer,
it assumes it is a libSBML return status code. If the code value is
LIBSBML_OPERATION_SUCCESS, returns without further action; if it is not,
prints an error message constructed using 'message' along with text from
libSBML explaining the meaning of the code, and exits with status code 1.
"""
if value == None:
raise SystemExit('LibSBML returned a null value trying to ' +
message + '.')
elif type(value) is int:
if value == libsbml.LIBSBML_OPERATION_SUCCESS:
return
else:
err_msg = 'Error encountered trying to ' + message + '.' \
+ 'LibSBML returned error code ' + str(value) + ': "' \
+ libsbml.OperationReturnValue_toString(value).strip() + '"'
raise SystemExit(err_msg)
else:
return
from biocrnpyler.sbmlutil import _create_global_parameter, _create_local_parameter
from biocrnpyler.sbmlutil import _create_modifiers, _create_products, _create_reactants
# generate an sbml model
document, model = create_sbml_model()
# If the document/model creation failed the following two lines
# will catch it:
check(document, 'create SBMLDocument object')
check(model, 'create SBML Model object')
S1 = Species("S1")
S2 = Species("S2")
compartment = model.getCompartment(0)
check(compartment, 'get compartment in SBML model')
sbml_species = add_species(model,
compartment,
S1,
initial_concentration=10)
sbml_species2 = add_species(model,
compartment,
S2,
initial_concentration=10)
check(sbml_species.getId(), 'get ID for SBML species')
check(sbml_species.getInitialConcentration(),
'get concentration for SBML species')
prop_hill = HillPositive(k=1, s1=S2, K=10, n=2)
crn_rxn = Reaction([S1], [], propensity_type=prop_hill)
check(model.createReaction(), 'create a new SBML reaction')
rx1 = model.getReaction(0)
check(rx1, 'get the created reaction')
check(rx1.setId('r1'), 'set ID for reaction')
check(rx1.setReversible(False), 'set reversible attribute for reaction')
_create_reactants(crn_rxn.inputs, rx1, model)
_create_products(crn_rxn.outputs, rx1, model)
_create_modifiers(crn_rxn, rx1, model)
check(rx1.getModifier(0), 'create species modifier')
modifier = rx1.getModifier(0)
check(modifier.getSpecies(),
'get the species id for the modifier reference')
rateLaw = rx1.createKineticLaw()
check(rateLaw, 'create kineticLaw for reaction')
local_param = _create_local_parameter(rateLaw, 'k_local', 10)
check(local_param, 'create local parameter in SBML model')
global_param = _create_global_parameter(model, 'k_global', 10)
check(global_param, 'create global parameter in SBML model')
check(rateLaw.setFormula('k_local*10 - k_global/2'),
'set rate formula for reaction')
validator = validateSBML(ucheck=False)
validation_result = validator.validate(document, print_results=True)
if validation_result > 0:
raise Exception('Invalid SBML model.')
def test_add_reaction_for_bioscrape():
"""
Generates models for bioscrape including the particular annotations needed.
"""
S1, S2, S3 = Species("S1"), Species("S2"), Species("S3")
species = [S1, S2, S3]
for_bioscrape = True #Toggle for_bioscrape
propensities = [
MassAction(k_forward=1, k_reverse=.1),
HillPositive(k=1, n=2., K=3., s1=S1),
HillNegative(k=1, n=2., K=3., s1=S1),
ProportionalHillPositive(k=1, n=2, K=3., s1=S1, d=S2),
ProportionalHillNegative(k=1, n=2, K=3., s1=S1, d=S2)
]
for prop in propensities:
rxn_num = 0
model_id = f"{prop.name}_model_with_for_bioscrape_{for_bioscrape}"
document, model = create_sbml_model(model_id=model_id)
add_all_species(model, species)
rxn = Reaction([S1], [S2, S3],
propensity_type=prop) #create a reaction
try:
sbml_reaction = add_reaction(
model, rxn, f"r{rxn_num}",
for_bioscrape=for_bioscrape) #add reaction to SBML Model
rxn_num += 1 #increment reaction id
crn_reactants = [str(S1)]
crn_products = [str(S2), str(S3)]
sbml_products = [
p.getSpecies() for p in sbml_reaction.getListOfProducts()
]
sbml_reactants = [
p.getSpecies() for p in sbml_reaction.getListOfReactants()
]
#test that the reaction has the right inputs and outputs
assert set(crn_reactants) == set(sbml_reactants)
assert set(crn_products) == set(sbml_products)
if len(crn_reactants) > 0:
assert all(
crn_reactants.count(s.getSpecies()) == int(
s.getStoichiometry())
for s in sbml_reaction.getListOfReactants())
if len(crn_products) > 0:
assert all(
crn_products.count(s.getSpecies()) == int(
s.getStoichiometry())
for s in sbml_reaction.getListOfProducts())
assert not sbml_reaction.getReversible()
# Check annotations:
sbml_annotation = sbml_reaction.getAnnotationString()
check_var = f"type={prop.name}" in str(sbml_annotation)
assert check_var == True
#Test that the sbml annotation has keys for all species and parameters
for k in prop.propensity_dict["parameters"]:
#convert k_reverse and k_forward to just k
k = k.replace("_reverse", "").replace("_forward", "")
check_var = f"{k}=" in sbml_annotation
assert check_var == True
#be sure that "k=" only shows up once in the annotation
assert sbml_annotation.count(f"{k}=") == 1
for s in prop.propensity_dict["species"]:
check_var = f"{s}=" in sbml_annotation
assert check_var == True
#TODO is there a smart way to test that the rate formula is correct?
assert validate_sbml(document) == 0 # Validate the SBML model
except Exception as e: #collect errors
error_txt = f"Unexpected Error: in sbmlutil.add_reaction {rxn} for {model_id}. \n {str(e)}."
raise Exception(error_txt)
# Names of different supported propensities
for prop in Propensity.get_available_propensities():
print(prop)
kb = 100
ku = 10
kex = 1.
kd = .1
G = Species(name="G", material_type="dna") #DNA
A = Species(name="A", material_type="protein") #Activator
GA = ComplexSpecies([G, A, A]) #Activated Gene
X = Species(name="X", material_type="protein")
rxnd = Reaction.from_massaction(inputs=[X], outputs=[], k_forward=kd)
# Massaction Unregulated
species1 = [G, A, GA, X]
rxn0_1 = Reaction.from_massaction(inputs=[G, A, A],
outputs=[GA],
k_forward=kb,
k_reverse=ku)
rxn0_2 = Reaction.from_massaction(inputs=[GA], outputs=[GA, X], k_forward=kex)
CRN0 = ChemicalReactionNetwork(species1, [rxn0_1, rxn0_2, rxnd])
mak1 = MassAction(k_forward=kb, k_reverse=ku)
mak2 = MassAction(k_forward=kex)
rxn1_1 = Reaction([G, A, A], [GA], propensity_type=mak1)
rxn1_2 = Reaction([G], [G, X], propensity_type=mak2)
