def test_optimize_turnover_metabolite(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(
milp_model, mutually_exclusive_directionality_constraint=False)
epsilon = 1.01 * solver_tolerance
gim3e.add_turnover_metabolites(milp_model, ['glc__D_c'], epsilon)
for x in milp_model.reactions:
x.objective_coefficient = 0
milp_model.reactions.get_by_id(
'TMS_glc__D_c').objective_coefficient = 1
gim3e.gim3e_optimize(milp_model,
objective_sense='maximize',
the_problem=None,
solver=the_solver,
error_reporting=None,
tolerance_optimality=solver_tolerance,
tolerance_feasibility=solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer=tolerance_integer)
solution = milp_model.solution
self.assertEqual(solution.status, "optimal")
def test_convert_objective_to_constraint(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(
milp_model, mutually_exclusive_directionality_constraint=True)
constraint_test_model = gim3e.convert_objective_to_constraint(
milp_model,
objective_sense='maximize',
fraction_of_optimum=0.9,
copy_model=True,
bound_best_optimum=False,
new_reaction_name="test_objective",
solver=the_solver,
tolerance_optimality=solver_tolerance,
tolerance_feasibility=solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer=tolerance_integer)
constraint_test_model.reactions.get_by_id(
"test_objective").objective_coefficient = 1.
gim3e.gim3e_optimize(milp_model,
objective_sense='maximize',
the_problem=None,
solver=the_solver,
error_reporting=None,
tolerance_optimality=solver_tolerance,
tolerance_feasibility=solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer=tolerance_integer)
solution = milp_model.solution
self.assertEqual(solution.status, "optimal")
def test_irreversible_flux_variability_analysis(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(
milp_model, mutually_exclusive_directionality_constraint=True)
the_test_reaction_id = "biomass_iRR1083_metals"
fva_dict = gim3e.irreversible_flux_variability_analysis(
milp_model,
fraction_of_optimum=0.9,
objective_sense='maximize',
the_reactions=[the_test_reaction_id],
solver=the_solver,
tolerance_optimality=solver_tolerance,
tolerance_feasibility=solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer=tolerance_integer,
error_reporting=None,
number_of_processes=1,
verbose=False)
self.assertEqual(fva_dict[the_test_reaction_id]['maximum_status'],
"optimal")
self.assertEqual(fva_dict[the_test_reaction_id]['minimum_status'],
"optimal")
def test_convert_objective_to_constraint(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(milp_model, mutually_exclusive_directionality_constraint = True)
constraint_test_model = gim3e.convert_objective_to_constraint(milp_model,
objective_sense = 'maximize',
fraction_of_optimum = 0.9,
copy_model = True,
bound_best_optimum = False,
new_reaction_name = "test_objective",
solver=the_solver,
tolerance_optimality = solver_tolerance,
tolerance_feasibility = solver_tolerance,
tolerance_barrier = 0.0001 * solver_tolerance,
tolerance_integer = tolerance_integer)
constraint_test_model.reactions.get_by_id("test_objective").objective_coefficient = 1.
gim3e.gim3e_optimize(milp_model, objective_sense = 'maximize',
the_problem=None, solver=the_solver,
error_reporting=None,
tolerance_optimality = solver_tolerance,
tolerance_feasibility = solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer = tolerance_integer)
solution = milp_model.solution
self.assertEqual(solution.status, "optimal")
def test_convert_to_irreversible(self):
from cobra.core.Reaction import Reaction
reaction_id_to_check = [x.id for x in self.model.reactions]
metabolite_id_to_check = [x.id for x in self.model.metabolites]
gene_id_to_check = [x.id for x in self.model.genes]
if cobra_version == '0.2.0':
reversible_model = deepcopy(self.model)
else:
reversible_model = self.model.copy()
# Since we won't solve for turnover eteabolites set mutually_exclusive_directionality_constraint = False
gim3e.convert_to_irreversible_with_indicators(
reversible_model,
mutually_exclusive_directionality_constraint=False)
for the_reaction_id in reaction_id_to_check:
the_reaction = reversible_model.reactions.get_by_id(
the_reaction_id)
if 'reflection' in dir(the_reaction):
if type(the_reaction.reflection) == Reaction:
the_reflection = the_reaction.reflection
# Check reactions point to the genes
gene_from_reaction = [
x.id for x in the_reaction.get_gene()
]
gene_from_reaction.sort()
gene_from_reflection = [
x.id for x in the_reflection.get_gene()
]
gene_from_reflection.sort()
self.assertEqual(gene_from_reaction, gene_from_reflection)
# Check that reactions point to the metabolites
metabolite_from_reaction = list(
the_reaction.get_products() +
the_reaction.get_reactants())
metabolite_from_reaction.sort()
metabolite_from_reflection = list(
the_reflection.get_products() +
the_reflection.get_reactants())
metabolite_from_reflection.sort()
self.assertEqual(metabolite_from_reaction,
metabolite_from_reflection)
for the_metabolite_id in metabolite_id_to_check:
the_metabolite = reversible_model.metabolites.get_by_id(
the_metabolite_id)
the_reaction_id_list = [
x.id for x in the_metabolite.get_reaction()
]
# Make sure forward and reverse reactions are both referenced
for the_reaction_id in the_reaction_id_list:
the_reaction = reversible_model.reactions.get_by_id(
the_reaction_id)
if 'reflection' in dir(the_reaction):
if type(the_reaction.reflection) == Reaction:
the_reflection = the_reaction.reflection
self.assertTrue(
the_reflection.id in the_reaction_id_list)
def test_solve_milp_feasible(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(milp_model, mutually_exclusive_directionality_constraint = True)
gim3e.gim3e_optimize(milp_model,objective_sense = 'maximize',
the_problem=None, solver=the_solver,
error_reporting=None,
tolerance_optimality = solver_tolerance,
tolerance_feasibility = solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer = tolerance_integer)
solution = milp_model.solution
self.assertEqual(solution.status, "optimal")
def test_evaluate_penalties(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(milp_model, mutually_exclusive_directionality_constraint = True)
expression_dict = {x.id: 1. for x in self.model.genes}
expression_dict.pop('s0001')
penalties = gim3e.evaluate_penalties(self.model, milp_model, expression_dict, 2)
for the_key in penalties.keys():
if len(milp_model.reactions.get_by_id(the_key).gene_reaction_rule) > 0:
if 's0001' not in (milp_model.reactions.get_by_id(the_key).gene_reaction_rule):
self.assertEqual(penalties[the_key], 1.)
else:
self.assertEqual(penalties[the_key], 0.)
else:
self.assertEqual(penalties[the_key], 0.)
def test_solve_milp_feasible(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(
milp_model, mutually_exclusive_directionality_constraint=True)
gim3e.gim3e_optimize(milp_model,
objective_sense='maximize',
the_problem=None,
solver=the_solver,
error_reporting=None,
tolerance_optimality=solver_tolerance,
tolerance_feasibility=solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer=tolerance_integer)
solution = milp_model.solution
self.assertEqual(solution.status, "optimal")
def test_optimize_turnover_metabolite(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(milp_model, mutually_exclusive_directionality_constraint = False)
epsilon = 1.01 * solver_tolerance
gim3e.add_turnover_metabolites(milp_model, ['glc__D_c'], epsilon)
for x in milp_model.reactions:
x.objective_coefficient = 0
milp_model.reactions.get_by_id('TMS_glc__D_c').objective_coefficient = 1
gim3e.gim3e_optimize(milp_model, objective_sense = 'maximize',
the_problem=None, solver=the_solver,
error_reporting=None,
tolerance_optimality = solver_tolerance,
tolerance_feasibility = solver_tolerance,
tolerance_barrier=0.0001 * solver_tolerance,
tolerance_integer = tolerance_integer)
solution = milp_model.solution
self.assertEqual(solution.status, "optimal")
def test_convert_to_irreversible(self):
from cobra.core.Reaction import Reaction
reaction_id_to_check = [x.id for x in self.model.reactions]
metabolite_id_to_check = [x.id for x in self.model.metabolites]
gene_id_to_check = [x.id for x in self.model.genes]
if cobra_version == '0.2.0':
reversible_model = deepcopy(self.model)
else:
reversible_model = self.model.copy()
# Since we won't solve for turnover eteabolites set mutually_exclusive_directionality_constraint = False
gim3e.convert_to_irreversible_with_indicators(reversible_model, mutually_exclusive_directionality_constraint = False)
for the_reaction_id in reaction_id_to_check:
the_reaction = reversible_model.reactions.get_by_id(the_reaction_id)
if 'reflection' in dir(the_reaction):
if type(the_reaction.reflection) == Reaction:
the_reflection = the_reaction.reflection
# Check reactions point to the genes
gene_from_reaction = [x.id for x in the_reaction.get_gene()]
gene_from_reaction.sort()
gene_from_reflection = [x.id for x in the_reflection.get_gene()]
gene_from_reflection.sort()
self.assertEqual(gene_from_reaction, gene_from_reflection)
# Check that reactions point to the metabolites
metabolite_from_reaction = list(the_reaction.get_products() + the_reaction.get_reactants())
metabolite_from_reaction.sort()
metabolite_from_reflection = list(the_reflection.get_products() + the_reflection.get_reactants())
metabolite_from_reflection.sort()
self.assertEqual(metabolite_from_reaction, metabolite_from_reflection)
for the_metabolite_id in metabolite_id_to_check:
the_metabolite = reversible_model.metabolites.get_by_id(the_metabolite_id)
the_reaction_id_list = [x.id for x in the_metabolite.get_reaction()]
# Make sure forward and reverse reactions are both referenced
for the_reaction_id in the_reaction_id_list:
the_reaction = reversible_model.reactions.get_by_id(the_reaction_id)
if 'reflection' in dir(the_reaction):
if type(the_reaction.reflection) == Reaction:
the_reflection = the_reaction.reflection
self.assertTrue(the_reflection.id in the_reaction_id_list)
def test_evaluate_penalties(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(
milp_model, mutually_exclusive_directionality_constraint=True)
expression_dict = {x.id: 1. for x in self.model.genes}
expression_dict.pop('s0001')
penalties = gim3e.evaluate_penalties(self.model, milp_model,
expression_dict, 2)
for the_key in penalties.keys():
if len(milp_model.reactions.get_by_id(
the_key).gene_reaction_rule) > 0:
if 's0001' not in (milp_model.reactions.get_by_id(
the_key).gene_reaction_rule):
self.assertEqual(penalties[the_key], 1.)
else:
self.assertEqual(penalties[the_key], 0.)
else:
self.assertEqual(penalties[the_key], 0.)
def test_irreversible_flux_variability_analysis(self):
the_solver = gim3e.check_solver()
tolerance_integer = gim3e.integer_tolerances[the_solver]
if cobra_version == '0.2.0':
milp_model = deepcopy(self.model)
else:
milp_model = self.model.copy()
gim3e.convert_to_irreversible_with_indicators(milp_model, mutually_exclusive_directionality_constraint = True)
the_test_reaction_id = "biomass_iRR1083_metals"
fva_dict = gim3e.irreversible_flux_variability_analysis(milp_model,
fraction_of_optimum = 0.9,
objective_sense= 'maximize',
the_reactions = [the_test_reaction_id],
solver = the_solver,
tolerance_optimality = solver_tolerance,
tolerance_feasibility = solver_tolerance,
tolerance_barrier = 0.0001*solver_tolerance,
tolerance_integer = tolerance_integer,
error_reporting = None,
number_of_processes = 1,
verbose = False)
self.assertEqual(fva_dict[the_test_reaction_id]['maximum_status'], "optimal")
self.assertEqual(fva_dict[the_test_reaction_id]['minimum_status'], "optimal")
