Task-based help
Five reproducible Evidence Atlas workflows
Each workflow ends in a stable page or downloadable object. Evidence levels are never treatment grades.
1 · How do I evaluate one gene?
- Open the ZNF133 research record.
- Read current action, legacy integrated DMD prior, source agreement, source state, prior-variant stability and evidence modules as separate fields.
- Inspect the evidence ladder and primary gap.
- Download the typed record for citation or reuse.
2 · How do I compare a candidate list?
Open the frozen example for ZNF133, MON1A and GFOD2. It has a stable object ID, checksum, rendered results, JSON and TSV that work without JavaScript. The interactive Compare form remains a temporary workspace until its output is preserved as an object.
3 · How do I interpret DMD source conflict?
Use DMD Evidence Explorer. “Conflicted” means at least one positive and one negative source-level median. Module membership and nine-variant stability do not replace this field.
4 · How do I generate and cite a Study Card?
Open the ZNF133 muscle-assay card. Cite its gene, card schema, resource release, evidence freeze and interface build; download YAML for preregistration or JSON/TSV for a registry.
5 · What does L2 not mean?
L2 is same-HepG2 observed perturbation or internal model support. It does not mean muscle replication, DMD functional validation, therapeutic rescue, safety or clinical benefit. L3a is an external-context screen; L3b is independent replication of the same perturbation.
Compact glossary
- L3a
- External-context screen; not same-perturbation replication.
- L3b
- Same perturbation replicated independently.
- L4
- DMD-relevant muscle functional validation.
- L5
- Therapeutic or clinical utility.
- Counteralignment
- Opposition to the integrated DMD prior; never therapeutic rescue.
- Not assessed
- Missing coverage, not a negative result.