Skip to content

Workflow CLI reference

This reference lists actual Pixi tasks and Typer wrapper commands present in the repository. The CLI remains the source of truth for workflow execution; the dashboard builds and previews these commands.

Config precedence

For workflows that support --conf, the dashboard currently validates assigned config files as TOML because the target runners parse TOML:

Explicit CLI flags override values from --conf.
Values from --conf override repository defaults.
If --conf is not supplied, repository config.toml is used.

Do not assume a custom config was used unless metadata.json and config_resolved.* confirm it.

Pixi tasks

Task Purpose Backend command/module Example
prep Preprocessing cleanup. python -m processing.cleanup pixi run prep
kinopt-local Local KinOpt. python -m kinopt.local pixi run kinopt-local --outdir results/example_kinopt
kinopt-evol Evolutionary KinOpt. python -m kinopt.evol pixi run kinopt-evol
kinopt-fitanalysis KinOpt fit analysis. python -m kinopt.fitanalysis pixi run kinopt-fitanalysis
tfopt-local Local TFOpt. python -m tfopt.local pixi run tfopt-local --outdir results/example_tfopt
tfopt-evol Evolutionary TFOpt. python -m tfopt.evol pixi run tfopt-evol
model ProtWise model runner. python -m protwise.runner.main pixi run model --conf config.toml --outdir results/example_protwise
networkmodel Global network model runner. python -m networkmodel.runner pixi run networkmodel --conf config.toml --output-dir results/example_networkmodel
phoskintime Typer workflow wrapper. python -m phoskintime.config.cli pixi run phoskintime --help
phoskintime-all Typer wrapper all command. python -m phoskintime.config.cli all pixi run phoskintime-all
network-dashboard Legacy Networkmodel dashboard launcher. python run_dashboard.py pixi run network-dashboard
dashboard Unified no-code dashboard. streamlit run dashboard/app.py pixi run dashboard
dashboard-dev Unified dashboard with reload-on-save. streamlit run dashboard/app.py --server.runOnSave=true pixi run dashboard-dev
test Run pytest with project path. PYTHONPATH=.:.. pytest pixi run -e dev test
test-notebooks Execute notebooks as tests. pytest --nbmake notebooks/*.ipynb pixi run -e dev test-notebooks
test-cov Coverage test run. pytest --cov=... pixi run -e dev test-cov
check-import Import smoke check. Python one-liner pixi run check-import
docs-serve Serve documentation. zensical serve pixi run -e docs docs-serve
docs-build Build documentation. zensical build pixi run -e docs docs-build

Typer wrapper commands (config/cli.py)

These commands are available through the package CLI wrapper when invoked as the configured module.

prep

  • Purpose: run preprocessing cleanup.
  • Backend module: processing.cleanup.
  • Required inputs: project preprocessing inputs.
  • Config support: no explicit --conf on wrapper command.
  • Output directory flag: none.
  • Example:
pixi run phoskintime prep

Dashboard integration: registered as prep.

kinopt

  • Purpose: run KinOpt local or evolutionary mode.
  • Backend module: kinopt.<mode>.
  • Required inputs: KinOpt inputs configured in config.toml/backend constants.
  • Config support: --conf.
  • Output directory flag: --outdir / --output-dir for local mode.
  • Optional arguments: --mode local|evol.
  • Example:
pixi run phoskintime kinopt --mode local --conf config.toml --outdir results/example_kinopt

Dashboard integration: local mode registered as kinopt-local.

tfopt

  • Purpose: run TFOpt local or evolutionary mode.
  • Backend module: tfopt.<mode>.
  • Required inputs: TFOpt inputs configured in config.toml/backend constants.
  • Config support: --conf.
  • Output directory flag: --outdir / --output-dir for local mode.
  • Optional arguments: --mode local|evol.
  • Example:
pixi run phoskintime tfopt --mode local --conf config.toml --outdir results/example_tfopt

Dashboard integration: local mode registered as tfopt-local.

model

  • Purpose: run ProtWise protein-wise ODE fitting.
  • Backend module: protwise.runner.main.
  • Required inputs: protein CSV, phosphosite Excel workbook, RNA/TFOpt Excel workbook.
  • Config support: --conf; custom config values are used unless overridden by explicit CLI fields.
  • Output directory flag: --outdir / --output-dir.
  • Example:
pixi run phoskintime model --conf config.toml --outdir results/example_protwise

Dashboard integration: registered as protwise-model / Pixi task model.

networkmodel

  • Purpose: run the integrated global network model.
  • Backend module: networkmodel.runner.
  • Required inputs: kinase network CSV, TF network CSV, MS/protein CSV, RNA CSV, optional phosphoproteomics CSV, optional KinOpt/TFOpt Excel priors.
  • Config support: --conf; custom config values are used unless overridden by explicit CLI fields.
  • Output directory flag: --outdir / --output-dir; dashboard uses --output-dir.
  • Example:
pixi run phoskintime networkmodel --conf config.toml --outdir results/example_networkmodel

Dashboard integration: registered as networkmodel.

all

  • Purpose: run preprocessing, TFOpt, KinOpt, and ProtWise model stages in sequence.
  • Backend modules: processing.cleanup, tfopt.<mode>, kinopt.<mode>, protwise.runner.main.
  • Config support: --tf-conf, --kin-conf, --model-conf.
  • Output directory flag: --outdir / --output-dir as a base directory for stage outputs.
  • Example:
pixi run phoskintime-all --outdir results/example_all

Dashboard integration: registered as phoskintime-all.

clean

  • Purpose: remove bytecode, cache, and build artifacts.
  • Backend: filesystem cleanup in config/cli.py.
  • Config support: none.
  • Output directory flag: none.
  • Example:
pixi run phoskintime clean

Dashboard integration: not a dashboard workflow.

Workflow details

Preprocessing / prep

  • Command name: prep.
  • Purpose: cleanup/preprocessing.
  • Backend module: processing.cleanup.
  • Required inputs: repository raw/processing inputs.
  • Optional arguments: none in the Pixi task.
  • Config file support: not exposed by task.
  • Output directory flag: none.
  • Expected outputs: preprocessing outputs defined by processing.cleanup.
  • Dashboard status: executable from dashboard registry; result browsing is limited unless outputs follow the result contract.

KinOpt local

  • Command name: kinopt-local.
  • Purpose: local kinase/phosphosite optimization.
  • Backend module: kinopt.local.
  • Required inputs: configured KinOpt input files, commonly CSV files under data/.
  • Optional arguments include --conf, --lower_bound, --upper_bound, --loss_type, and --method where supported.
  • Output directory flag: --outdir / --output-dir.
  • Expected outputs: kinopt_results.xlsx, tables, plots, reports, logs, provenance.
  • Dashboard status: executable and result-browsable.

KinOpt evol

  • Command name: kinopt-evol.
  • Purpose: evolutionary KinOpt optimization.
  • Backend module: kinopt.evol.
  • Inputs/config: backend-defined.
  • Output directory flag: not dashboard-standardized in the Pixi task.
  • Dashboard status: not directly registered for dashboard execution; produced folders can be browsed if layout is recognized.

TFOpt local

  • Command name: tfopt-local.
  • Purpose: local TF/mRNA optimization.
  • Backend module: tfopt.local.
  • Required inputs: configured TFOpt files, commonly CSV files under data/.
  • Optional arguments include --conf, --lower_bound, --upper_bound, and --loss_type where supported.
  • Output directory flag: --outdir / --output-dir.
  • Expected outputs: tfopt_results.xlsx, tables, plots, reports, logs, provenance.
  • Dashboard status: executable and result-browsable.

TFOpt evol

  • Command name: tfopt-evol.
  • Purpose: evolutionary TFOpt optimization.
  • Backend module: tfopt.evol.
  • Inputs/config: backend-defined.
  • Output directory flag: not dashboard-standardized in the Pixi task.
  • Dashboard status: not directly registered for dashboard execution; produced folders can be browsed if layout is recognized.

ProtWise / model

  • Command name: model.
  • Purpose: protein-wise ODE fitting.
  • Backend module: protwise.runner.main.
  • Required inputs: protein CSV, phosphosite Excel workbook, RNA/TFOpt Excel workbook.
  • Config support: --conf.
  • Output directory flag: --outdir / --output-dir.
  • Example:
pixi run model --conf config.toml --outdir results/example_protwise
  • Expected outputs: result workbook, plots, report, logs, provenance, standard folders.
  • Dashboard status: executable and result-browsable.

Networkmodel

  • Command name: networkmodel.
  • Purpose: global coupled signaling/GRN model.
  • Backend module: networkmodel.runner.
  • Required inputs: kinase network CSV, TF network CSV, MS/protein CSV, RNA CSV, optional phosphoproteomics CSV; optional prior workbooks from KinOpt/TFOpt.
  • Config support: --conf.
  • Output directory flag: --output-dir / --outdir.
  • Example:
pixi run networkmodel --conf config.toml --output-dir results/example_networkmodel
  • Expected outputs: scalar objective/convergence CSVs, predictions, dashboard_bundle.pkl, plots, optimization/profile/posterior outputs, logs, provenance.
  • Dashboard status: executable and result-browsable.

Dashboard

  • Commands: dashboard, dashboard-dev, network-dashboard.
  • Purpose: dashboard and dashboard-dev run the unified dashboard; network-dashboard runs the legacy Networkmodel dashboard.
  • Backend modules: dashboard/app.py or run_dashboard.py.
  • Example:
pixi run dashboard
pixi run dashboard-dev

Tests and docs

  • Test command:
pixi run -e dev test
  • Focused dashboard tests:
pixi run -e dev pytest tests/dashboard -v
  • Docs build:
pixi run -e docs docs-build

Standalone analysis scripts

The scripts/ directory contains standalone utilities, including:

  • analyze_tf_kin_counts.py
  • compare_mechanisms.py
  • curve_similarity.py
  • export_subnetworks.py
  • find_protein_accumulators.py
  • kinopt_network_readout.py
  • kinopt_network_viz.py
  • make_kinopt_diagram.py
  • mechanistic_insights.py
  • temporal_sensitivity.py
  • tfopt_network_readout.py
  • tfopt_network_viz.py

These are not all Pixi tasks. Run them directly only after reading each script's arguments and expected inputs. When integrating a script into the dashboard, parameterize input/output paths and write outputs into the selected run directory under tables/, plots/, reports/, or artifacts/.

File-format constraints

  • Do not document YAML/JSON execution config support unless the target runner implements YAML/JSON loading; dashboard --conf assignments are restricted to .toml.
  • Do not document TSV/XLSX execution support for Networkmodel network/data inputs unless the backend reader is extended beyond default pandas.read_csv.
  • Do not document Excel support for ProtWise protein input unless the backend reader changes from pd.read_csv.
  • Dashboard upload preview can read more formats than a specific workflow can execute.