3. Draft model generation from COBRA/ pyTFA

SKiMpy can generate kinetic draft models from pyTFA or cobrapy models, assigning best guess reaction mechanisms.

Cobrapy and pyTFA models can be used in generate unparametrized draft models:

from skimpy.io.generate_from_cobra import FromCobra
from cobra.io.mat import load_matlab_model

# Load a cobra model
this_cobra_model = load_matlab_model('/skimpy/models/toy_model.mat','model')

# Generate a draft kinetic model
model_gen = FromCobra()
kmodel = model_gen.import_model(this_cobra_model)

When building draft models based on cobrapy we assume full reversibility of all reactions using pyTFA models this can be specified by accesing the results of the Gibbs free energy vairables.

import pytfa
from pytfa.io import import_matlab_model, load_thermoDB
from pytfa.io.viz import get_reaction_data

# Convert the cobra model to a thermodynamics model
thermo_data = load_thermoDB('/skimpy/data/thermo_data.thermodb')
this_pytfa_model = pytfa.ThermoModel(thermo_data, this_cobra_model)

GLPK = 'optlang-glpk'
this_pytfa_model.solver = GLPK

## TFA conversion
this_pytfa_model.prepare()
this_pytfa_model.convert(add_displacement=True)

# Solve the pytfa model to find deltaGs
solution = this_pytfa_model.optimize()

# Generate the KineticModel

# Define the molecules that should be considered small-molecules
# These molecules will not be accounted explicitly in the kinetic mechanism as
# substrates and products
small_molecules = ['h_c', 'h_e']

model_gen = FromPyTFA(small_molecules=small_molecules)
kmodel = model_gen.import_model(this_pytfa_model,solution.raw)