Frequently Asked Questions

What is QualiBact?

QualiBact is a community-driven framework for species-specific quality control of bacterial genome assemblies. We curate thresholds, visualizations, and reference datasets to help labs adopt consistent QC standards.

Who maintains the thresholds?

Thresholds are derived from the QualiBact pipeline (typically the qualibact-v1.0 scheme) and reviewed by the CGPS team plus community contributors. Each release lists the scheme version used for every species.

How often is QualiBact updated?

We update the dataset whenever we finish a new curation cycle—usually every few months, or sooner when large data contributions arrive. Check the Summary page or release notes for the most recent timestamp.

How do I request a new species?

See the Organism Requests page for the current queue and submission instructions. If your species is missing, email [email protected] with relevant details so we can prioritize it.

Can I contribute my own thresholds or data?

Yes! The Contributing page outlines two workflows: submitting raw assembly metrics so we can run QualiBact, or sharing your own validated thresholds for inclusion. Contributors receive credit in releases and publications.

What formats do you support for downloads?

All tabular results are provided as CSV (often compressed as .csv.xz). Plots are PNG files. The Next.js app also includes manifests in JSON—such as refseq_metrics.json—to help downstream tooling discover assets.

For programmatic access without scraping the site, the /api/v2/ endpoints expose flat threshold tables:

/api/v2/thresholds.csv — one row per (species, scheme, metric)
/api/v2/thresholds.json — nested JSON
/api/v2/index.json — registry of species + schemes + preferred scheme

Can I query QualiBact from Claude / Cursor / other LLM clients?

Yes. QualiBact runs an MCP (Model Context Protocol) server at https://mcp.qualibact.org/mcp. Any MCP-aware client can call it directly instead of web-searching for thresholds.

To wire it into Claude Desktop, Claude Code, Cursor, etc., add one entry to your MCP config file (location varies by client — see your client's docs):

{
  "mcpServers": {
    "qualibact": {
      "transport": {
        "type": "http",
        "url": "https://mcp.qualibact.org/mcp"
      }
    }
  }
}

Reload the client and five tools become available:

Tool	What it returns
`list_species(genus?)`	All species + preferred scheme + every scheme that has data
`list_schemes()`	Every QC scheme + species count
`get_thresholds(species, scheme?, metric?)`	The 4-bound (FAIL_lower / WARN_lower / WARN_upper / FAIL_upper) rows
`list_pinned_thresholds(species?)`	All hand-pinned expert overrides with rationale
`get_pinned_threshold(species, scheme, metric)`	Rationale for one specific pin

Once it's wired in, asking "What's the QualiBact N50 floor for E. coli?" gets you the authoritative answer (FAIL_lower = 20000, WARN_lower = 23000) instead of an educated guess.

The MCP is read-only, public (no auth), and read-through-caches the same /api/v2/ blobs above.

Where can I report issues or bugs?

Open an issue on the QualiBact GitHub repository with as much context as possible (species, scheme, error logs). You can also email the maintainers if the problem involves sensitive data.

NCBI's GenBank genome-size-check publishes daily-updated expected genome-size ranges per species (taxid → min / max / expected) at https://ftp.ncbi.nlm.nih.gov/genomes/ASSEMBLY_REPORTS/species_genome_size.txt.gz. It covers ~6,000 taxa and is a good independent sanity check. QualiBact's Genome_Size thresholds are typically tighter than NCBI's min–max range (we reject partial/incomplete assemblies that NCBI's range accommodates); for the species the two share, 95% of QualiBact bounds are within ±50% of NCBI's, and zero have both sides drift >50%.

For read-level QC (sequencing depth, contamination identity, heterozygosity) see bactscout as the recommended companion tool — QualiBact operates on the assembled genome.

Funding

QualiBact is supported by the National Institute for Health Research (NIHR) under grant number NIHR133307. The views expressed are those of the maintainers and contributors, and do not necessarily reflect those of the NIHR or the UK Department of Health and Social Care.