Skip to content

refactor: take experiments, solver support, MW and SMILES from raven-toolbox - #98

Draft
edkerk wants to merge 4 commits into
developfrom
feat/experiment-data
Draft

edkerk wants to merge 4 commits into
developfrom
feat/experiment-data

Conversation

@edkerk

@edkerk edkerk commented Nov 20, 2026 •

Copy link
Copy Markdown
Collaborator

geckopy takes the functions that need no enzymes from raven-toolbox (CompMetEng/raven-toolbox#204).

This is a draft. It needs a release of raven-toolbox with these functions first, and the raven-toolbox requirement in pyproject.toml has to name that release.

  • Experiments: FluxData, load_flux_data, load_experiment_data, simulate_conditions, experiment_distance and compute_carbon_weights are in raven_toolbox.experiments, and geckopy.databases still exports FluxData and load_flux_data. apply_flux_data_constraints calls the function of the same name in raven-toolbox and takes the biomass reaction and the carbon source from the model adapter.
  • Names of the experiment data: the evotune functions take experiment_data and max_growth, FluxData.weight is the weight of a condition, and the files are fluxData.tsv, maxGrowth.tsv and zeroFlux.tsv. load_experiment_data takes the data folder instead of the adapter. The column evotuneRMSEweight raises an error that names weight. In the tutorial fluxData.tsv holds the condition of the protein step as its first row, with weight 0, so the tuning leaves it out.
  • Solver: geckopy.solver_support is removed; its functions are in raven_toolbox.utils.solver. EcModel still repairs the scipy problem whenever model.solver is read.
  • Molecular weight and SMILES: calculate_mw, find_met_smiles and met_smiles are in raven-toolbox, and geckopy.databases exports them from there. find_met_smiles needs cache_path.

The MATLAB side is CompMetEng/GECKO#38 and CompMetEng/RAVEN#28.

…tune

The loader, the condition simulation and the distance live in
geckopy.experiments, where the kcat tuning and the flux constraints of the
protein step can both use them. The data files are fluxData.tsv,
maxGrowth.tsv and zeroFlux.tsv, and the weight column is called weight.
The tutorial's fluxData.tsv holds the single protein-step condition with
weight 0 next to the tuning conditions.
…en-toolbox

FluxData, load_flux_data, load_experiment_data, simulate_conditions,
experiment_distance and compute_carbon_weights are in raven_toolbox.experiments,
and apply_flux_data_constraints calls the function of the same name there and
adds the biomass reaction and the carbon source of the model adapter.
load_experiment_data takes the data folder instead of the adapter.
The functions of geckopy.solver_support are in raven_toolbox.utils.solver. EcModel
still repairs the scipy problem whenever model.solver is read.
They are in raven_toolbox.utils.protein and raven_toolbox.annotation, and
geckopy.databases exports them from there. find_met_smiles needs cache_path.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant