Skip to content

Rewrite the guide for GECKO 4: a quick path and decision pages - #44

Open
edkerk wants to merge 24 commits into
developfrom
doc/gecko4-guide
Open

edkerk wants to merge 24 commits into
developfrom
doc/gecko4-guide

Conversation

@edkerk

@edkerk edkerk commented Nov 12, 2026 •

Copy link
Copy Markdown
Collaborator
  • The pull request is for main.

Main changes

The GECKO and geckopy submodules move to the merged code through the daily submodule update; this pull request waits for that.

Pages that changed

  • New: guide/first-ecmodel, guide/ecmodel-structure, guide/kcat-sources, guide/custom-kcats, guide/protein-pool, guide/kcat-tuning, guide/advanced-tuning, guide/proteomics, guide/relaxing-proteomics, guide/analysis, guide/engineering-targets, guide/context-ecmodel, guide/enzyme-usage.
  • Rewritten: guide/index, guide/introduction.
  • Removed (no redirects): guide/getting-started, guide/workflow, guide/building-ec-model, guide/gathering-kcats, guide/applying-kcats, guide/gecko-light, guide/growth-rate-tuning, guide/tuning-against-experimental-data, guide/proteomics-integration, guide/relaxing-constraints, guide/simulation-and-analysis, guide/enzyme-usage-and-bottlenecks, and the workflow figures.
  • Updated: gecko3-to-gecko4, gecko-to-geckopy, gecko4-kcat-assignment, installation, guide/basic-model-operations, guide/data-folder-files, guide/model-adapter-parameters, guide/yaml-format, index.

Choose a kcat tuning now compares sensitivityTuning and evotune and keeps the
checks that apply to both. The standard tuning, with the fit of sigma after
it, moves to Tune kcats with sensitivityTuning. The advanced tuning page is
now Tune kcats with evotune and states that sigma is not fitted after it.
Build an ecModel, the protein pool page and the workflow diagrams name both
tunings.
A kcat tuning raises kcats and cannot lower a growth rate that ends above the
target; sigmaFitter then lowers sigma. A growth rate below the target is
corrected through the kcats, not through sigma or the pool. Keeps the text of
the GECKO 3 protocol paper in reference/ as a source for the documentation.
Scripts for both toolboxes on the full_ecModel tutorial that rank the
kcats, tune with two seeds and list the large changes, plus the solver,
a Slurm job, the number of kcats and generations, a check against
left-out conditions and how to carry corrections into custom.tsv.
The context page explains what differs on that route, and the overview
has a path from the context page back to Build an ecModel.
…oups and BRENDA aggregation

[kcat_sources.okp] methods names the OpenKineticsPredictor methods and [okp] holds the request options. Evotune source groups are one table each with sources and sigma0_log, with prefix patterns and the default groups custom, brenda, ec_tree and combined. [kcat_sources.brenda] aggregation sets the BRENDA value of direct calls, the median for the default preset. gr_exp has no default, unknown keys are errors, and showAdapterParams and geckopy adapter show print the resolved parameters. The GECKO 3 to GECKO 4 page states the median default and the adapter class fields that are not carried over.
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