E. coli group 2 and group 3 capsular K-typing database

Database update summary: v4.0.0

Database v4.0.0 expands the group 2 and group 3 reference set from 90 to 109 K-locus records. It adds 16 sequence-defined loci, assigned the new EC-K-Typing identifiers KL176–KL191 following comparison with the kTYPr capsule catalogue ^[1]. These loci do not currently have assigned K phenotypes. A crosswalk between the EC-K-Typing identifiers, kTYPr catalogue labels, and source genome accessions is provided in the v4 locus provenance. The update also adds three references with known phenotype associations: K18-K22 (KL18), K94 (KL94), and K95 (KL95).

Three existing locus records have been renamed to match their established K phenotypes: the records previously designated KL124, KL110, and KL116 are now designated KL6, KL51, and KL97, respectively. K74 is excluded because it is no longer recognised as a valid K serotype.

KL158 now has the descriptive type Capsule null predicted. When comparing results generated with a database version before v4.0.0, we recommend rerunning Kaptive with the current database or applying the nomenclature crosswalk above.

About the database

This Kaptive database supports in silico typing of E. coli group 2 and group 3 capsular K loci, which use ABC transporter-dependent capsule assembly systems ^[2].

The reference set was developed by screening approximately 50,000 E. coli genomes from bloodstream infection, human and animal carriage, and collections spanning Europe, North America, Africa, and Asia. This was supplemented with group 2 kpsF-positive and group 3 kpsM-positive assemblies identified at 90% k-mer identity in a collection of 661,000 bacterial assemblies ^[3–13].

The database is expected to cover most group 2 and group 3 loci found among invasive E. coli. These capsule groups are especially prevalent in phylogroups B2 and D and occur less frequently in phylogroups A, B1, and C.

The database includes all currently recognised group 2 and group 3 K phenotypes with an established locus association: 33 K phenotypes represented by 30 locus records. Three phenotype pairs cannot currently be distinguished reliably from their K-locus sequences and are reported as composite types: K13-K23 (KL13), K18-K22 (KL18), and K96-K54 (KL96). KL158 additionally has the descriptive type Capsule null predicted.

Disrupted loci and phenotype interpretation

Some reference loci are closely related to loci with established K phenotypes but contain one or more disrupted genes. Disruption of a capsule biosynthesis or modification gene may prevent capsule production or alter the polysaccharide sufficiently that the parental K phenotype is no longer appropriate. A disrupted locus is therefore not assigned the phenotype of its nearest intact relative unless that association is independently supported.

These sequence-defined variants are retained as separate KL records so that Kaptive can report the observed genotype directly. If they were removed, an isolate carrying the disrupted locus could instead match the nearest intact reference and receive an unsupported phenotype assignment. When interpreting an individual query assembly, any additional gene disruptions or missing genes should be considered alongside assembly quality. A Kaptive Typeable result describes confidence in the locus match, not proof of capsule production.

Atypical loci and reference boundaries

Some K loci have atypical architectures in which capsule-specific genes occur outside the conserved export regions. A current list of loci classified as having atypical architecture is provided in the atypical-locus metadata.

A close Kaptive match may fail to capture additional capsule genes located beyond the reference boundaries. For these loci, inspect the assembly context and gene-level Kaptive results. See the Kaptive documentation for general guidance on interpreting results.

Database construction

Reference genes were annotated using Bakta 1.10.x and reconciled using Panaroo 1.5.2. The final reference GFFs were generated using Panaroo 1.6.0 ^[14,15].

Non-capsular IS-element-associated annotations and newly assigned hypothetical CDS annotations shorter than 200 bp were excluded from the final gene annotations. Their underlying nucleotide sequences were not removed from the locus references.

The reproducible database-construction workflow and curated reference GFFs are provided in DB/panaroo_refset.

Database additions and corrections

To have a novel locus considered for inclusion in the public database, contact rebeccgl@uio.no or open an issue in this repository.

Requests to add or correct a K-phenotype association for an existing locus are also welcome. Please provide the locus designation, proposed phenotype, reference strain or sequence accession where available, and the supporting serological, experimental, or published evidence. Phenotype associations will only be added when their provenance and relationship to the locus can be evaluated.

For sequences that cannot yet be shared, the containerised private database builder can add one local locus to a frozen copy of the public reference set and produce a validated private Kaptive database. The sequence is processed locally, the repository is not modified, and nothing is uploaded. Local identifiers of KL9000 or greater are deliberately kept separate from official EC-K-Typing locus assignments.

Using the database

The database is formatted for Kaptive 3 and has been validated using Kaptive 3.1.0. After cloning the repository or downloading a release, run:

kaptive assembly EC-K-typing_group2and3.gbk your_assembly.fasta

See the Kaptive documentation for installation, output interpretation, and additional options.

Citation

If you use this database, please cite:

Gladstone, R. A., Pesonen, M., Pöntinen, A. K. et al. Identification of transporter-dependent capsular loci associated with the invasive potential of Escherichia coli. Nature Microbiology 11, 1205–1216 (2026). https://doi.org/10.1038/s41564-026-02283-w

If you use Kaptive, please also cite:

Stanton, T. D., Hetland, M. A. K., Löhr, I. H., Holt, K. E. & Wyres, K. L. Fast and accurate in silico antigen typing with Kaptive 3. Microbial Genomics 11, 001428 (2025). https://doi.org/10.1099/mgen.0.001428

References

1. Miravet-Verde S, Cacace E, Mores CR, Rutschmann C, Lin C-W, et al. In silico typing maps the natural diversity of Escherichia coli transporter-dependent capsules. Nature Microbiology. 2026;11:1217–1232.
2. Stanton TD, Hetland MAK, Löhr IH, Holt KE, Wyres KL. Fast and accurate in silico antigen typing with Kaptive 3. Microbial Genomics. 2025;11:001428.
3. Gladstone RA, McNally A, Pöntinen AK, Tonkin-Hill G, Lees JA, et al. Emergence and dissemination of antimicrobial resistance in Escherichia coli causing bloodstream infections in Norway in 2002–17: a nationwide, longitudinal, microbial population genomic study. The Lancet Microbe. 2021;2:e331–e341.
4. Arredondo-Alonso S, Pöntinen AK, Gama JA, Gladstone RA, Harms K, et al. Plasmid-driven strategies for clone success in Escherichia coli. Nature Communications. 2025;16:2921.
5. Kallonen T, Brodrick HJ, Harris SR, Corander J, Brown NM, et al. Systematic longitudinal survey of invasive Escherichia coli in England demonstrates a stable population structure only transiently disturbed by the emergence of ST131. Genome Research. 2017;27:1437–1449.
6. Pöntinen AK, Gladstone RA, Pesonen H, Pesonen M, Cléon F, et al. Modulation of multidrug-resistant clone success in Escherichia coli populations: a longitudinal, multi-country, genomic and antibiotic usage cohort study. The Lancet Microbe. 2024;5:e142–e150.
7. Shao Y, Garcia-Mauriño C, Clare S, Dawson NJR, Mu A, et al. Primary succession of Bifidobacteria drives pathogen resistance in neonatal microbiota assembly. Nature Microbiology. 2024;9:2570–2582.
8. Mäklin T, Thorpe HA, Pöntinen AK, Gladstone RA, Shao Y, et al. Strong pathogen competition in neonatal gut colonisation. Nature Communications. 2022;13:7417.
9. Liu CM, Aziz M, Park DE, Wu Z, Stegger M, et al. Using source-associated mobile genetic elements to identify zoonotic extraintestinal E. coli infections. One Health. 2023;16:100518.
10. Ludden C, Raven KE, Jamrozy D, Gouliouris T, Blane B, et al. One Health genomic surveillance of Escherichia coli demonstrates distinct lineages and mobile genetic elements in isolates from humans versus livestock. mBio. 2019;10:e02693-18.
11. Sands K, Carvalho MJ, Portal E, Thomson K, Dyer C, et al. Characterization of antimicrobial-resistant Gram-negative bacteria that cause neonatal sepsis in seven low- and middle-income countries. Nature Microbiology. 2021;6:512–523.
12. Dicks J, Fazal M-A, Oliver K, Grayson NE, Turnbull JD, et al. NCTC3000: a century of bacterial strain collecting leads to a rich genomic data resource. Microbial Genomics. 2023;9:mgen000976.
13. Blackwell GA, Hunt M, Malone KM, Lima L, Horesh G, et al. Exploring bacterial diversity via a curated and searchable snapshot of archived DNA sequences. PLoS Biology. 2021;19:e3001421.
14. Schwengers O, Jelonek L, Dieckmann MA, Beyvers S, Blom J, Goesmann A. Bakta: rapid and standardized annotation of bacterial genomes via alignment-free sequence identification. Microbial Genomics. 2021;7:000685.
15. Tonkin-Hill G, MacAlasdair N, Ruis C, Weimann A, Horesh G, et al. Producing polished prokaryotic pangenomes with the Panaroo pipeline. Genome Biology. 2020;21:180.