Kraken2/Bracken

What does it do?

Use Kraken2/Bracken to classify sequence data taxonomically and estimate the organisms represented in a sample.

The workflow runs:

  • Kraken2 — assigns sequencing reads or assembly sequences to taxonomic groups using a selected reference database;
  • Bracken — refines abundance estimates from Kraken2 classifications.

The automator supports paired-end raw reads, Nanopore reads, and assemblies. Database choice is important because the standard Kraken2 and PlusPF databases are intended for metagenomic analysis, while the Greengenes, RDP, and SILVA options are 16S-oriented databases.

For background, see the Kraken2 repository, the Kraken2 protocol, Wood and Salzberg (2014), and Wood et al. (2019). Cite the appropriate tool authors when publishing results produced with this automator.

How do I use it?

Subject

In the Subject field, enter:

kraken2

Spelling matters, but matching is not case-sensitive.

Description

Enter optional settings first, followed by one SEQID per line.

Minimal request using paired-end raw reads and the standard database:

2026-SEQ-0001
2026-SEQ-0002

Attachments

No attachment is required. The automator retrieves the sequence data associated with each requested SEQID according to seqtype.

Optional parameters

seqtype

Selects the input data type.

  • Default: paired
  • Paired-end raw reads: seqtype=paired
  • Nanopore reads: seqtype=nanopore
  • Assembly data: seqtype=assembly

database

Selects the taxonomic reference database.

  • Default: kraken2
  • Supported values:
  • database=kraken2
  • database=plusPF
  • database=greengenes
  • database=rdp
  • database=silva

The default kraken2 database is documented as the standard database dated September 26, 2022, obtained from the Kraken2 Index Zone.

Use kraken2 or plusPF for metagenomic classification. The greengenes, rdp, and silva options are 16S-based databases and should be selected only when appropriate for the input and analysis goal.

Examples

Paired-end metagenomic reads

database=kraken2
2026-SEQ-0001
2026-SEQ-0002

Nanopore metagenomic reads with PlusPF

seqtype=nanopore
database=plusPF
2026-MIN-0001

Assembly classification

seqtype=assembly
database=kraken2
2026-SEQ-0001

See issue 29355 for an example Kraken2 request. Temporary result-download links associated with the issue may expire.

Interpreting results

When Kraken2/Bracken finishes, it uploads an archive named using the Redmine issue identifier:

kraken2_output_redmineID.zip

The archive contains Kraken2 and Bracken report files.

Use Kraken2 output to review taxonomic classifications and Bracken output to review refined abundance estimates. Interpret low-abundance or closely related taxa cautiously because classification accuracy depends on sequence quality, database composition, and similarity among represented organisms.

Results produced with a 16S-oriented database are not directly equivalent to results produced with a metagenomic database. Record the selected database when reporting or comparing results.

How long does it take?

Runtime depends on the selected input type, database, amount of sequence data, and number of requested samples. Kraken2 is faster than the original Kraken workflow, but large requests can require substantial memory.

Limit metagenomic requests to approximately 10 samples at a time to reduce the risk of memory exhaustion.

What can go wrong?

A requested SEQID is unavailable

Symptom: The Redmine issue receives a warning identifying unavailable sequences.

Likely cause: The automator cannot locate the data required for the requested SEQID and seqtype.

What to do: Verify each SEQID, confirm that the selected input type exists, and submit a corrected request.

The request runs out of memory

Symptom: The job fails while processing a large metagenomic batch.

Likely cause: Too many samples or a memory-intensive database was requested at once.

What to do: Split the request into batches of no more than approximately 10 metagenomes and resubmit them separately.

The selected database is inappropriate

Symptom: Results are sparse, misleading, or difficult to compare with prior analyses.

Likely cause: A 16S-oriented database was selected for metagenomic data, or a metagenomic database was selected for an analysis intended to use a 16S reference collection.

What to do: Choose kraken2 or plusPF for metagenomic classification and use greengenes, rdp, or silva only for an appropriate 16S-based analysis.

The selected input type is incorrect

Symptom: The automator cannot locate or process the expected sequence files.

Likely cause: seqtype does not match the available paired-end reads, Nanopore reads, or assembly data.

What to do: Correct seqtype and resubmit the request.

  • MetaPhlAn4 — profiles microbial communities using marker genes and supports raw reads and assemblies.
  • AutoCLARK — reports species represented in raw reads or draft assemblies using CLARK.
  • StrainMash — compares an assembly with RefSeq type strains to identify its closest reference.