robustnet 0.2.0

1. Estimate the reference-state fluxes

The first step is to estimate a reference steady-state flux distribution consistent with the defined metabolic network using available fluxomics data (typically from ¹³C metabolic flux analysis).

model.load_priors(
  'fluxomics',
  data=flux_measurements,
  std=flux_measurement_uncertainties
)

fit_res = model.estimate_reference_fluxes(
  bounds=flux_bounds,
  exclude_metabolites=excluded_from_mass_balance
)

2. Generate ensemble model parameter sets

Next, posterior parameter distributions are inferred using multi-omics measurements and enzyme kinetic priors. The resulting sampled parameter sets are used to construct ensembles of kinetic models for downstream simulations.

model.load_priors(
  'reference_fluxes',
  data=fit_res.estimated_fluxes,
  std=fit_res.estimated_flux_errors
)
model.load_priors(
  'metabolomics',
  data=metab_measurements,
  std=metab_measurement_uncertainties
)
model.load_priors(
  'proteomics',
  data=enz_measurements,
  std=enz_measurement_uncertainties
)
model.load_priors(
  'kparameters',
  data=kparam_measurements,
  std=kparam_measurement_uncertainties
)

samp_res = model.generate_parameter_sets(
  alpha=pentality_strength,
  n_samples=n_samples_to_draw,
  n_jobs=n_parallel_jobs
)

The sampling procedure integrates information from experimental measurements and enzyme databases while maintaining consistency across heterogeneous data sources and mechanistic constraints.

3. Simulate metabolic responses to perturbations

Using the parameterized ensemble models, metabolic responses and robustness can be evaluated under specified enzyme perturbations.

model.load_parameter_sets(
  mconc_set=samp_res.sampled_metabolite_concentrations,
  econc_set=samp_res.sampled_enzyme_concentrations,
  kparam_set=samp_res.sampled_kinetic_parameters
)

rob_res = model.evaluate_robustness(
  perturb_enzymes=enz_list,
  fold_change=perturbation_amplitude,
  n_models=n_models_to_generate,
  n_jobs=n_parallel_jobs
)

This workflow enables multi-omics-informed analysis of metabolic robustness and system-level responses to enzyme perturbations. The package also provides helper utilities and visualization functions for inspecting model parameterization, perturbation trajectories, robustness metrics, and sensitivity analyses.

For more detailed tutorials and API references, please refer to the documentation.