Condition Spec

Condition specs define what was tested and under which context. Well-defined specs enable reproducibility, comparison, and collaboration.

Why it matters

Results depend heavily on context. Even the same target can yield different outcomes depending on disease, cell type, or treatment conditions.

Required fields

  • dataset_id: dataset or model identifier
  • mode: simulation type (e.g., knockdown)
  • targets: genes/proteins to perturb
  • condition: condition label

Recommended metadata

  • Baseline name (control, untreated, etc.)
  • Disease or cohort (ALS, NSCLC, responder, etc.)
  • Cell or tissue type
  • Timepoint
  • Treatment/culture conditions
  • Readouts (markers, rankings, summaries)

Example

{
  "dataset_id": "ds_als_motor_neuron_v1",
  "mode": "knockdown",
  "targets": ["GENE_A", "GENE_B"],
  "condition": "sporadic_motor_neuron_d18"
}