The interpretation hierarchy
Four layers, one consistent flow
Each layer builds on the one before it. The flow is the same for every ecosystem and every report family.
Input features
Sequencing-derived signals from the sample — taxonomic, ecological and (where supported) functional. The raw evidence layer.
Scores
Conservative, transparent summaries of those features. Each score states its method support and a confidence cue. Never a diagnosis.
Panels
Biological-context containers that group related scores under a theme — the unit of interpretation.
Insight Packs
The primary report layer. Packs group panels under a theme; some unlock with method, package, QC, repeat testing or professional context.
Score tiers
Depth scales with the sequencing method
Higher tiers become available only where the chosen method and data quality support them — never assumed, never inflated.
Foundation scores
Baseline composition, diversity and ecological-context scores. Broadly method-supported.
Core intelligence scores
Deeper taxonomic and ecological structure with higher resolution where the method supports it.
Advanced scores
Pathway-potential and expression-context outputs — only where the sequencing tier and data quality support them.
What unlocks deeper layers
Panels and packs are gated, not guessed
Additional panels and Insight Packs become available based on a few clear conditions. Eligibility is decided at runtime during configuration and reporting — never hardcoded into the report itself.
Reportable context only
The consumer-facing report shows reportable microbiome context only. ATLAS scores describe microbial composition, diversity and ecological context relative to reference data — they do not diagnose, treat, predict or cure disease, and should be read with symptoms, history and clinician assessment where relevant.
See the architecture in action on a real report
Explore how scores, panels and Insight Packs come together — then configure the depth you need.