Community profiling
Profile which microbes are present and how abundant they are across samples, starting from a processed feature table and an optional taxonomy map.
Research question
Which taxa dominate each sample and cohort, how prevalent are they, and how much of the community remains unclassified at the chosen rank?
Who this is for
- Microbiome researchers characterizing a new cohort or environment
- Core facilities producing standardized profiling and QC reports
- Teams onboarding processed 16S/ITS or shotgun profiles for faster analysis
Data requirements
| Data | Required | Purpose |
|---|---|---|
| Feature table (taxa × samples) | Yes | Abundance and prevalence computation |
| Taxonomy map (feature → lineage) | No | Collapse features to a rank and label taxa |
| Reference database + version | Recommended | Provenance for every taxonomic claim |
Workflow
Upload feature table (+ taxonomy) → QC and provenance
→ Taxonomic profiling at the chosen rank
→ Review top taxa, prevalence, and unclassified fraction
→ AI interpretation
Step 1 — QC and provenance
Run QC to review per-sample depth, feature retention, prevalence, and low-depth flags. The reference database and version are recorded on the run so conclusions stay reproducible.
Step 2 — Taxonomic profiling
Run the Taxonomy analysis to compute relative abundance, prevalence, and the top taxa at your chosen rank, with explicit handling of unclassified and ambiguous assignments. A shared rank selector (phylum through ASV) on the analyze page drives this and the diversity, core microbiome, and differential abundance steps consistently.
Step 3 — Review the stacked bar chart
A per-sample stacked bar chart shows the top 10 taxa (with an "Other" category) at the selected rank, so you can scan composition across the cohort before moving to formal comparisons.
Expected outputs
- Per-sample depth and feature-retention QC summary
- Top taxa with mean relative abundance and prevalence
- Per-sample stacked relative-abundance bar chart at the selected rank
- Mean unclassified fraction at the selected rank
- Reference-database provenance recorded on the run