Workflow overview#

The diagram shows the order of the guide pages, from a new project to the analysis of an ecModel, with the functions of each page. The boxes with a thick outline are the steps of Build an ecModel, which makes a working ecModel with the default settings; that page skips the functions marked optional, and the steps from the ecModel structure to the kcat tuning each have a decision page with the options. A solid arrow connects pages that run one after another, a dotted arrow is an optional path, a diamond a choice between paths, and a dashed outline an optional page. A click on a box opens its page. The button in the corner of a diagram opens it in full screen, where the mouse wheel zooms and a drag pans. In the page, the wheel zooms with Ctrl (Cmd on macOS) held. The download icon at the top of each tab saves its diagram as a PDF.

        ---
config:
  flowchart:
    useMaxWidth: false

---
flowchart TB
  classDef quick stroke-width:3px
  classDef opt stroke-dasharray: 5 4
  P["<b>Build an ecModel</b><br/>createGECKOproject<br/>fill model_adapter.toml"]:::quick
  S["<b>Choose the ecModel structure</b><br/>ModelAdapterManager.setDefault<br/>loadConventionalGEM<br/>makeEcModel (full or light)<br/>getComplexData, applyComplexData (optional)"]:::quick
  K["<b>Choose the kcat sources</b><br/>getECfromGEM, then getECfromDatabase<br/>findMetSmiles<br/>updateKcatSources<br/>collectKcats (optional preview)<br/>assignKcats"]:::quick
  C["<b>Set custom kcats</b><br/>data/kcats/custom.tsv<br/>convertCustomKcats (required for a<br/>GECKO 3 data/customKcats.tsv)"]:::opt
  PP["<b>Set the protein pool</b><br/>setProtPoolSize<br/>calculateFfactor (optional)"]:::quick
  G{"Measured data"}
  T["<b>Choose a kcat tuning</b><br/>standard: sensitivityTuning<br/>(no change when growth reaches the target)<br/>then sigmaFitter"]:::quick
  A["<b>Tune kcats against experimental data</b><br/>screenKcatLeverage<br/>reviewKcatAssignment (optional)<br/>cmaesKcatTuning (selects with selectTunableMask)"]
  F["<b>Functional ecModel</b><br/>saveEcModel"]:::quick
  P --> S --> K --> PP --> G
  C -.-> K
  T -.->|keep a tuned kcat| C
  G -->|one maximum growth rate| T
  G -->|measured fluxes or several growth rates| A
  T --> F
  A --> F
  PR["<b>Constrain enzymes with proteomics</b> (full ecModel)<br/>loadProtData, fillEnzConcs<br/>levelComplexConcs (optional), constrainEnzConcs<br/>getIsozymeCoverage, zeroUnmeasuredIsozymes (optional)<br/>loadFluxData, calculateFfactor (optional)<br/>setProtPoolSize, constrainFluxData"]:::opt
  F -.-> PR
  F -.-> CX
  CX -.->|proteomics of the context| PR
  CX -.-> RD
  RL{"Growth reaches the measured growth rate?"}
  PR --> RL
  RX["<b>Relax proteomics constraints</b><br/>flexibilizeEnzConcs (recommended),<br/>relaxProteomicsGreedy when enzymes at 0 limit,<br/>relaxEnzConcs to compare"]
  RL -->|no| RX
  RD["ecModel for simulation and analysis"]
  RL -->|yes| RD
  RX --> RD
  F -.->|without proteomics| RD
  AN["<b>Analyse an ecModel</b><br/>ecFVA, plotEcFVA<br/>randomSampling (RAVEN)<br/>mapRxnsToConv"]
  EN["<b>Find engineering targets</b><br/>ecFSEOF"]
  CX["<b>Make a context-specific ecModel</b><br/>getSubsetEcModel, before proteomics"]:::opt
  EU["<b>Enzyme usage and bottlenecks</b> (full ecModel)<br/>getEnzymeUsage, reportEnzymeUsage<br/>getEnzymeBottlenecks, getConcControlCoeffs<br/>getPfbaEnzymes"]
  RD --> AN
  RD --> EN
  RD --> EU
  click P href "first-ecmodel.html"
  click S href "ecmodel-structure.html"
  click K href "kcat-sources.html"
  click C href "custom-kcats.html"
  click PP href "protein-pool.html"
  click T href "kcat-tuning.html"
  click A href "advanced-tuning.html"
  click F href "first-ecmodel.html"
  click PR href "proteomics.html"
  click RX href "relaxing-proteomics.html"
  click AN href "analysis.html"
  click EN href "engineering-targets.html"
  click CX href "context-ecmodel.html"
  click EU href "enzyme-usage.html"
    
        ---
config:
  flowchart:
    useMaxWidth: false

---
flowchart TB
  classDef quick stroke-width:3px
  classDef opt stroke-dasharray: 5 4
  P["<b>Build an ecModel</b><br/>geckopy init<br/>fill model_adapter.toml"]:::quick
  S["<b>Choose the ecModel structure</b><br/>ModelAdapter.from_folder<br/>load_conventional_gem<br/>make_ec_model (full or light)<br/>get_complex_data, apply_complex_data (optional)"]:::quick
  K["<b>Choose the kcat sources</b><br/>load_uniprot_tsv, load_kegg_tsv<br/>fill_eccodes_from_gem, then fill_eccodes_from_database<br/>find_met_smiles<br/>update_kcat_sources<br/>collect_kcats (optional preview)<br/>assign_kcats"]:::quick
  C["<b>Set custom kcats</b><br/>data/kcats/custom.tsv<br/>convert_custom_kcats (required for a<br/>GECKO 3 data/customKcats.tsv)"]:::opt
  PP["<b>Set the protein pool</b><br/>set_prot_pool_size<br/>calculate_f_factor (optional)"]:::quick
  G{"Measured data"}
  T["<b>Choose a kcat tuning</b><br/>standard: sensitivity_tuning<br/>(no change when growth reaches the target)<br/>then fit_sigma"]:::quick
  A["<b>Tune kcats against experimental data</b><br/>screen_kcat_leverage<br/>review_assignment (optional)<br/>cmaes_kcat_tuning (selects with select_tunable_mask)"]
  F["<b>Functional ecModel</b><br/>save_ec_model"]:::quick
  P --> S --> K --> PP --> G
  C -.-> K
  T -.->|keep a tuned kcat| C
  G -->|one maximum growth rate| T
  G -->|measured fluxes or several growth rates| A
  T --> F
  A --> F
  PR["<b>Constrain enzymes with proteomics</b> (full ecModel)<br/>load_prot_data, fill_enz_concs<br/>level_complex_concs (optional), constrain_enz_concs<br/>get_isozyme_coverage, zero_unmeasured_isozymes (optional)<br/>load_flux_data, calculate_f_factor (optional)<br/>set_prot_pool_size, apply_flux_data_constraints"]:::opt
  F -.-> PR
  F -.-> CX
  CX -.->|proteomics of the context| PR
  CX -.-> RD
  RL{"Growth reaches the measured growth rate?"}
  PR --> RL
  RX["<b>Relax proteomics constraints</b><br/>flexibilize_enz_concs (recommended),<br/>relax_proteomics_greedy when enzymes at 0 limit,<br/>relax_enz_concs to compare"]
  RL -->|no| RX
  RD["ecModel for simulation and analysis"]
  RL -->|yes| RD
  RX --> RD
  F -.->|without proteomics| RD
  AN["<b>Analyse an ecModel</b><br/>ec_fva<br/>random_sampling (raven-toolbox)<br/>map_rxns_to_conv"]
  EN["<b>Find engineering targets</b><br/>ec_fseof"]
  CX["<b>Make a context-specific ecModel</b><br/>get_subset_ec_model, before proteomics"]:::opt
  EU["<b>Enzyme usage and bottlenecks</b> (full ecModel)<br/>enzyme_usage, report_enzyme_usage<br/>get_enzyme_bottlenecks, get_conc_control_coeffs<br/>pfba_enzymes"]
  RD --> AN
  RD --> EN
  RD --> EU
  click P href "first-ecmodel.html"
  click S href "ecmodel-structure.html"
  click K href "kcat-sources.html"
  click C href "custom-kcats.html"
  click PP href "protein-pool.html"
  click T href "kcat-tuning.html"
  click A href "advanced-tuning.html"
  click F href "first-ecmodel.html"
  click PR href "proteomics.html"
  click RX href "relaxing-proteomics.html"
  click AN href "analysis.html"
  click EN href "engineering-targets.html"
  click CX href "context-ecmodel.html"
  click EU href "enzyme-usage.html"
    

Pages, order and alternatives#

Each row is a guide page with the page it runs after and what it leaves behind. The functions of each page are listed in the diagram and in the Functions table of the page.

Page

Run after

Output

Build an ecModel

none

a project folder with model_adapter.toml, and at the end a saved functional ecModel

Choose the ecModel structure

project folder

ecModel with enzymes and protein pool, without kcats

Choose the kcat sources

ecModel structure

data/kcats/<source>.tsv, and ec.kcat applied as enzyme constraints

Set custom kcats (optional)

the ecModel structure, or a kcat tuning whose result is kept; read by the first stage of every preset

data/kcats/custom.tsv

Set the protein pool

applied kcats

upper bound of the protein pool exchange

Choose a kcat tuning

protein pool; the standard tuning changes no kcat when the growth rate reaches the target

tuned ec.kcat, and \(\sigma\) fitted to the growth rate

Tune kcats against experimental data

protein pool, and measured maximum growth rates, exchange fluxes or both, of one or more conditions; alternative to the standard tuning

tuned ec.kcat

Constrain enzymes with proteomics (optional)

functional full ecModel

proteomics-constrained ecModel

Relax proteomics constraints

proteomics-constrained ecModel below the measured growth rate

relaxed ecModel

Analyse an ecModel, Find engineering targets, Enzyme usage and bottlenecks

a functional, proteomics-constrained or relaxed ecModel (enzyme usage: full ecModel only); independent of each other

results

Make a context-specific ecModel (optional)

functional ecModel without proteomics constraints

a context-specific ecModel, to which the proteomics of the context can be added

Reading the paths#

  • Quick path: Build an ecModel runs the boxes with a thick outline in one go, without the functions marked optional and with the default of each decision page. A decision page is needed only to change a default.

  • Kcats: updateKcatSources (Python: update_kcat_sources) makes the source files that the preset of model_adapter.toml reads, and assignKcats (Python: assign_kcats) selects one kcat per reaction from them, with the custom kcats first.

  • Tuning: the standard tuning raises kcats one at a time until one target growth rate is reached, and sigmaFitter (Python: fit_sigma) then fits \(\sigma\) to that growth rate. Tuning against experimental data fits many kcats to measured growth rates, exchange fluxes or both, of one or more conditions.

  • Relaxation: flexibilizeEnzConcs is the method to start with; relaxProteomicsGreedy raises enzymes with a bound of 0, and relaxEnzConcs gives a result that does not depend on the order in which enzymes become limiting, to compare with (Relax proteomics constraints).

  • Analysis: the analysis pages do not depend on each other. Enzyme usage and proteomics need a full ecModel. A context-specific ecModel is trimmed from an ecModel without proteomics constraints, and the proteomics of the context are added afterwards.

See also#