BioUnfold.in
All protocols

Download a reference genome with NCBI datasets

Pull any reference genome assembly from NCBI by accession or taxon with the datasets CLI, then unzip it to a ready-to-use FASTA.

SSSudipta SardarUpdated July 21, 2026

Prerequisites

  • NCBI datasets CLI installed (`conda install -c conda-forge ncbi-datasets-cli`)
  • An assembly accession or taxon name

1Download the assembly package

Ask datasets for a single assembly by its accession and restrict the package to the genome sequence with --include genome. The tool bundles everything into ncbi_dataset.zip in the current directory.

bash
datasets download genome accession GCF_000005845.2 --include genome

Expected output

Collecting 1 genome record[s] [================================================] 100% 1/1
Downloading: ncbi_dataset.zip 1.35MB done

2Unzip and locate the FASTA

Extract the archive into a named directory so multiple downloads never collide, then find the genomic FASTA nested under ncbi_dataset/data. The .fna file is the reference sequence you feed to downstream tools.

bash
unzip ncbi_dataset.zip -d ecoli
find ecoli -name '*.fna'

Expected output

Archive:  ncbi_dataset.zip
  inflating: ecoli/README.md
  inflating: ecoli/ncbi_dataset/data/assembly_data_report.jsonl
  inflating: ecoli/ncbi_dataset/data/GCF_000005845.2/GCF_000005845.2_ASM584v2_genomic.fna
  inflating: ecoli/ncbi_dataset/data/dataset_catalog.json
ecoli/ncbi_dataset/data/GCF_000005845.2/GCF_000005845.2_ASM584v2_genomic.fna

Troubleshooting

datasets: command not found

The CLI is not installed or not on your PATH. Install it from the conda-forge channel: `conda install -c conda-forge ncbi-datasets-cli`.

You have a species name, not an accession

Query by taxon and let NCBI pick the reference assembly: `datasets download genome taxon "Escherichia coli" --reference --include genome`.

References