Host-pathogen integration
Group episode outcomes by a pathogen or host feature carried in versioned cross-area results, preserving the originating area, run, and version.
Research question
Do outcomes differ by pathogen lineage, resistance, or a host feature produced in another area — without recomputing that area's science here?
Who this is for
- Translational infection researchers relating pathogen features to outcomes
- Teams combining Microbiology lineage/AMR results with clinical episodes
- Analysts who need provenance retained across linked evidence
Data requirements
| Data | Required | Purpose |
|---|---|---|
| Episodes with outcomes | Yes | The outcome to group and compare |
| Linked results referencing another area | Yes | Feature (e.g. lineage) grouping by episode |
| Feature key in the linked result summary | Recommended | Which summary field to group by |
Workflow
Link versioned Microbiology/CompBio/Immunology results by episode
→ Run integration with a feature key (e.g. lineage)
→ Review outcomes grouped by feature, with provenance
→ AI interpretation
Step 1 — Link results
Record cross-area results against the correct episode and specimen. Each link keeps the area, run id, method, and version — the integration contract is shared subject/episode/specimen identity, not a copied pipeline.
Step 2 — Integrate
The Integration analysis groups episode outcomes by a feature (such as lineage) from the linked result summaries and reports each group's event rate together with its linked provenance.
Expected outputs
- Outcome groups by pathogen/host feature with event rates
- Linked provenance table (area, run, version, result type)
- Counts of episodes and linked results integrated