BacDive — Bacterial Diversity Metadatabase (DSMZ)

It is the reference card catalog of the microbial world, compiled by the Leibniz Institute DSMZ - an ELIXIR Core Data Resource and a Global Core Biodata Resource - and standardized to a degree few collections match: 102,187 bacterial and archaeal strain records, including 22,126 type strains, each carrying its own citable Digital Object Identifier such as 10.13145/bacdive1.20260601.11. Every record is organized into ten observed sections: general identification; name and taxonomic classification backed by LPSN down to species rank, synonyms included; morphology; culture and growth conditions; physiology and metabolism; isolation, sampling and environmental origin; interaction and safety; sequence information; genome-based predictions; and literature.

What separates this record from most of the shelf is that the standardization survives delivery. Colony color, shape and incubation period arrive as fields; culture conditions arrive as a named medium, a temperature range with optimum and a pH range; safety arrives as a biosafety level with comment sitting beside recorded pathogenicity toward human, animal and plant hosts. Datadory ships those ten sections as one typed table keyed on BacDive-ID.

The record pools into Datadory's life & health insurance data hub as one of the slice's 16 primary datasets. Get a sample of this dataset cut to the taxa, habitats or safety classes you actually screen.

One record per strain, straight from the August 2026 verification pass:

# delivered grain: one record per strain, sectioned by data category

BacDive-ID          : 132485
taxonomy            : Lysobacter tolerans
description         : aerobe, Gram-negative, rod-shaped bacterium that forms circular colonies
culture_medium      : LB (Luria-Bertani) MEDIUM
temperature_optimum : 28 C
isolation           : soil from hexachlorocyclohexane-contaminated dumpsite - India, Asia

# a second strain, same row shape
BacDive-ID      : 1      DSM-Number   : 2002
taxonomy        : Acetobacter aceti    NCBI tax id  : 435
biosafety_level : 1      doi          : 10.13145/bacdive1.20260601.11

Read the first row as three answers already decided for you: how the organism is classified (Lysobacter tolerans, aerobe, Gram-negative rods), how to grow it (LB medium, optimum 28 C), and where it came from (soil at a hexachlorocyclohexane-contaminated dumpsite in India, Asia). The second row shows the identifier spine every record rides: a DSMZ culture-collection number, an NCBI tax id, a biosafety level and a per-strain DOI on one line, so a citation, a join and a safety review all resolve without opening a browser tab.

Twelve documented fields define the core record, each definition verified against live output during the August 2026 research pass - a bar met by 85.7 percent of the 1,744 datasets Datadory catalogs. They split into five jobs:

• Identity: BacDive-ID, DSM-Number, the full scientific name with its LPSN-backed lineage and synonyms, the NCBI tax id and the per-strain DOI - the columns that make joins, citations and reclassification tracking possible.
• Husbandry: recommended culture medium with whether growth occurs, plus temperature and pH ranges with their type.
• Origin: isolation sample type with country, continent and latitude-longitude.
• Handling risk: biosafety level with comment under the German risk-group classification, and recorded pathogenicity findings for human, animal and plant hosts.
• Cross-reference: INSDC 16S rRNA and genome accession numbers with GC content.

The wider collection runs past any single page - fatty acid profiles, antibiotic susceptibility discs, metabolite utilization, spore formation and halophily all exist in the physiology section - so anything outside the twelve-column core ships as additional fields on request, documented in the same dictionary format once your sample names them.

Three chips summarize the footprint:

• Geography: global - strains isolated worldwide, with country, continent and geolocation carried as fields rather than left in prose.
• Temporal: a continuously curated collection; numbered releases carry milestone dates from December 2023 through December 2025, and the 2025 edition crossed the 100,000-strain mark while adding a genome browser.
• Granularity: one record per strain (102,187), sectioned by data category.

Context for the collectors: this record scores 9/10 on Datadory's quality rubric against a catalogue mean of 7.81 - a tier shared by 534 of the 1,744 datasets catalogued - and it is one of the 16 primary datasets pooled in a slice averaging 8.31.

API, files, or your warehouse. Daily, weekly, or hourly.

Pick the channel and set the cadence to match the decision being fed; the rows arrive identical either way - cleaned, typed and keyed on BacDive-ID so the taxonomy spine, the culture panel and the safety panel line up without merge-key archaeology. Section structure resolves onto one stable record shape, multi-value physiology stays parseable, and every column is checked against the dictionary above before it reaches you. A sample scoped to your taxa, habitats or safety cuts ships first either way, with the field dictionary pinned against the delivered records.

Six jobs the strain ledger settles outright:

1. Strain selection and safety triage - biosafety level and pathogenicity columns cut 102,187 organisms down to biosafety-level-1 candidates before anyone designs an experiment.
2. Bioprospecting by habitat - isolation-source categories with geography turn 'which environments yield alkaliphiles' into a filter instead of a literature review.
3. Culture-condition lookup - medium, temperature window and pH optimum for a hard-to-grow isolate, delivered as fields rather than footnote forensics.
4. Taxonomic verification - LPSN-backed names with synonyms and NCBI tax ids keep a reclassification from breaking a pipeline mid-quarter.
5. Feature engineering - morphology, growth ranges and habitat flags as model training columns; see ml model training.
6. Citation-grade sourcing - the per-strain DOI gives a headline claim a stable identifier; see citation-grade research.

Market sizing on the tools and services that surround microbial work draws on the same collection; see market sizing.

Data Scientists & ML Engineers (relevance 3 of 3) get 102,187 typed strain records with morphology, growth ranges and habitat flags ready for feature work - see data scientists in life & health insurance. Developers & Data-Product Builders (3 of 3) stand up organism lookup and screening products on a stable BacDive-ID grain - see developers builders use cases. Journalists, Academics & Students (3 of 3) cite per-strain DOIs rather than someone's chart of a database - see journalists academics use cases. Market Researchers & Consultants (2 of 3) map microbial-diversity research activity by geography and habitat - see market researchers use cases.

The edge worth naming: this is descriptive metadata about organisms, not measured biochemistry. Kinetic values live in BRENDA Enzyme Database, and the two are compared directly in vs BRENDA Enzyme Database.

Methodology note - records are community deposits curated at the Leibniz Institute DSMZ, so sections populate only where deposition data exist: one strain may carry fatty acid profiles while another stops at taxonomy and safety. Depth gets checked per record before anything ships rather than assumed uniform.

Completeness note - the twelve-column core travels on every record; section-level extras ship as additional fields on request, pinned with examples once a sample names the families it needs.

Provenance note - every record traces to a per-strain Digital Object Identifier, type-strain status where applicable, and INSDC sequence accessions - a chain of custody short enough to verify by eye.

Where to go next:

• BRENDA Enzyme Database - the measured half of the lab bench: kinetic parameters across roughly 8,800 enzyme classes (how it compares).
• ATCC — Biological Materials Catalog - the acquisition-side counterpart: authenticated materials with catalog numbers, applications and pricing.
• EBI BioSamples — Biological Sample Metadata Database - specimen-to-assay linkage when the unit shifts from organism to sample.
• Addgene Plasmid Repository Catalog - the construct ledger that pairs with organism selection in any engineering workflow.
• DSMZ BacDive source profile - the curator behind the collection.

Browse the rest of the shelf on the life & health insurance data hub or the best life sciences tools & services datasets ranking, and read why a verified data dictionary matters downstream.