DFAST PlasmidFinder Additional Database Integration Summary

Generated: 2026-06-11  |  DFAST ver. 1.4.0  |  PlasmidFinder v3 (blastn)  |  3 additional DBs (distributed via figshare)

Three additional databases for replicon typing and plasmid typing have been integrated into DFAST's PlasmidFinder reference data. In addition to the standard PlasmidFinder database (enterobacteriales, Inc18, Rep1–3, RepA_N, etc.), the following are automatically fetched and registered by dfast_file_downloader.py --plasmidfinder and used for plasmid (replicon type) typing during an --amr run.

1. List of additional databases

DB nameDescriptionFile SequencesSizeSource (figshare)
PseudomonasRepDBPseudomonas replicon typing database v1.0 PseudomonasRepDB_v1.fsa404421 KB PseudomonasRepDB database v1.0
WHRepDBWHRepDB database v1.0 WHRepDB_v1.fsa12,92312 MB WHRepDB database v1.0
AcinetobacterAcinetobacterPlasmidTyping database v3 AcinetobacterPlasmidTyping_v3.fsa555561 KB AcinetobacterPlasmidTyping database v3.0

Sizes are the values after installation into db/plasmidfinder_db/; sequence counts are the number of > lines in the FASTA. The figshare file ID, file name, and size of each dataset were confirmed via the figshare API to match the download URL (ndownloader.figshare.com/files/<id>).

2. FASTA header format

All are distributed in PlasmidFinder v3 format (underscore-delimited, containing no spaces) and can be used directly as blastn subjects. No prior indexing (makeblastdb / KMA) is required.

DB nameExample headerNotes
PseudomonasRepDB>PInc-1alpha__CP002151.1
WHRepDB>WH-A1__NZCP038616.1
Acinetobacter>R3-T1_1__NC010605.1The original headers contain spaces, so they are automatically corrected on download (spaces → underscores)

3. Usage

Substitute <DB_ROOT> with your DFAST database directory (e.g. dfast_core/db).

Step 1: Fetch and register the reference data

Cloning of the standard PlasmidFinder DB and fetching/registration of the 3 additional DBs from figshare are performed automatically.

$ dfast_file_downloader.py --plasmidfinder -d <DB_ROOT>

The data is downloaded to <DB_ROOT>/plasmidfinder_db, and rows for the additional DBs are appended to <DB_ROOT>/plasmidfinder_db/config.

Re-running is safe. This command deletes and recreates the plasmidfinder_db directory every time, so the additional DBs are never registered more than once no matter how often it is run.

Step 2: Run DFAST

Plasmid (replicon type) typing is enabled with the --amr option (which includes PlasmidFinder). The additional DBs are used automatically.

$ dfast --genome your_genome.fna --amr -o OUTPUT_DIR

Step 3: Check the results

Matched replicon types are written to the contig (source feature) annotation and to amr_summary.tsv. The 3 fetched additional DBs are each registered in <DB_ROOT>/plasmidfinder_db/config and used for typing.

Record of sources: the figshare dataset page for each DB (linked in the table above) is also recorded as a comment in each entry of plasmidfinder_extra_databases in the source scripts/dfast_file_downloader.py. Direct downloads use URLs of the form ndownloader.figshare.com/files/<id> (the figshare.com/... form of the dataset page returns an empty response and cannot be used).