Download FASTQ files from the European Nucleotide Archive or the Sequence Read Archive repositories.
fastq-dl takes an ENA/SRA accession (Study, Sample, Experiment, or Run) and queries ENA (via
Data Warehouse API) to determine the associated
metadata. It then downloads FASTQ files for each Run. For Samples or Experiments with multiple
Runs, users can optionally merge the runs.
fastq-dl is available from Bioconda and I highly recommend you
go this route to for installation.
conda create -n fastq-dl -c conda-forge -c bioconda fastq-dl
conda activate fastq-dl
fastq-dl --help
Usage: fastq-dl [OPTIONS]
Download FASTQ files from ENA or SRA.
╭─ Required Options ──────────────────────────────────────────────────────────────────────────╮
│ * --accession -a TEXT ENA/SRA accession to query. (Study, Sample, Experiment, Run │
│ accession) │
│ [required] │
╰─────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Download Options ──────────────────────────────────────────────────────────────────────────╮
│ --provider [ena|sra] Specify which provider (ENA or SRA) to use. │
│ [default: ena] │
│ --group-by-experiment Group Runs by experiment accession. │
│ --group-by-sample Group Runs by sample accession. │
│ --max-attempts -m INTEGER Maximum number of download attempts. [default: 10] │
│ --sra-lite Set preference to SRA Lite │
│ --only-provider Only attempt download from specified provider. │
│ --only-download-metadata Skip FASTQ downloads, and retrieve only the │
│ metadata. │
│ --ignore -I Ignore MD5 checksums for downloaded files. │
╰─────────────────────────────────────────────────────────────────────────────────────────────╯
╭─ Additional Options ────────────────────────────────────────────────────────────────────────╮
│ --outdir -o TEXT Directory to output downloads to. [default: ./] │
│ --prefix TEXT Prefix to use for naming log files. [default: fastq] │
│ --cpus INTEGER Total cpus used for downloading from SRA. [default: 1] │
│ --force -F Overwrite existing files. │
│ --silent Only critical errors will be printed. │
│ --sleep -s INTEGER Minimum amount of time to sleep between retries (API query and │
│ download) │
│ [default: 10] │
│ --version -V Show the version and exit. │
│ --verbose -v Print debug related text. │
│ --help -h Show this message and exit. │
╰─────────────────────────────────────────────────────────────────────────────────────────────╯
fastq-dl requires a single ENA/SRA Study, Sample, Experiment, or Run accession and FASTQs
for all Runs that fall under the given accession will be downloaded. For example, if a Study
accession is given all Runs under that studies umbrella will be downloaded. By default,
fastq-dl will try to download from ENA first, then SRA.
The accession you would like to download associated FASTQS for. Currently the following types of accessions are accepted.
| Accession Type | Prefixes | Example |
|---|---|---|
| BioProject | PRJEB, PRJNA, PRJDB | PRJEB42779, PRJNA480016, PRJDB14838 |
| Study | ERP, DRP, SRP | ERP126685, DRP009283, SRP158268 |
| BioSample | SAMD, SAME, SAMN | SAMD00258402, SAMEA7997453, SAMN06479985 |
| Sample | ERS, DRS, SRS | ERS5684710, DRS259711, SRS2024210 |
| Experiment | ERX, DRX, SRX | ERX5050800, DRX406443, SRX4563689 |
| Run | ERR, DRR, SRR | ERR5260405, DRR421224, SRR7706354 |
The accessions are using regular expressions from the ENA Training Modules - Accession Numbers section.
fastq-dl gives you the option to download from ENA or SRA. the --provider option will
specify which provider you would like to attempt downloads from first. If a download fails
from the first provider, additional attempts will be made using the other provider.
ENA was selected as the default provider because the FASTQs are available directly without the need for conversion.
By default, fastq-dl will fallback on a secondary provider to attempt downloads. There
may be cases where you would prefer to disable this feature, and that is exactly the
purpose of --only-provider. When provided, if a FASTQ cannot be downloaded from the
original provider, no additional attempts will be made.
There maybe times you might want to group Run accessions based on a Experiment or Sample
accessions. This will merge FASTQs associated with a Run accession based its associated
Experiment accession (--group-by-experiment) or Sample accession (--group-by-sample).
Downloads from SRA are provided in SRA Normalized and SRA Lite formats.
SRA Normalized is the original format with full base quality scores and SRA Lite are smaller
due to simplifying the quality scores to a uniform Q30. By default the preference will be
set to SRA Normalized, if you prefer SRA Lite you can use --sra-lite to set the
preference to SRA Lite.
| Extension | Description |
|---|---|
-run-info.tsv |
Tab-delimited file containing metadata for each Run downloaded |
-run-mergers.tsv |
Tab-delimited file merge information from --group-by-experiment or --group-by-sample |
.fastq.gz |
FASTQ files downloaded from ENA or SRA |
Sometimes you might be reading a paper and they very kindly provided a BioProject of all
the samples they sequenced. So, you decide you want to download FASTQs for all the samples
associated with the BioProject. fastq-dl can help you with that!
fastq-dl --accession PRJNA248678 --provider SRA
fastq-dl --accession PRJNA248678
The above commands will download the 3 Runs that fall under Study accession PRJNA248678
from either SRA (--provider SRA) or ENA (without --provider).
Let's say instead of the whole BioProject you just want a single Experiment. You can do that as well.
fastq-dl --accession SRX477044
The above command would download the Run accessions from ENA that fall under Experiment SRX477044.
The relationship of Experiment to Run is a 1-to-many relationship, or there can be many Run accessions
associated with a single Experiment Accession (e.g. re-sequencing the same sample). Although in most
cases, it is a 1-to-1 relationship, you can use --group-by-experiment to merge multiple runs
associated with an Experiment accession into a single FASTQ file.
Ok, this time you just want a single Sample, or Biosample.
fastq-dl --accession SRS1904245 --provider SRA
The above command would download the Run accessions from SRA that fall under Sample SRS1904245.
Similar to Experiment accessions, the relationship of Sample to Run is a 1-to-many relationship,
or there can be many Run accessions associated with a single Sample Accession. Although in most
cases, it is a 1-to-1 relationship, you can use --group-by-sample to merge multiple runs
associated with an Sample accession into a single FASTQ file.
_Warning! For some type strains (e.g. S. aureus USA300) a Biosample accession might be associated with
100s or 1000s of Run accessions. These Runs are likely associated with many different conditions and
really should not fall under a single BioSample accession. Please consider this when using
--group-by-sample.
Let's keep it super simple and just download a Run.
fastq-dl --accession SRR1178105 --provider SRA
The above command would download the Run SRR1178105 from SRA. Run accessions are the end of the line (1-to-1 relationship), so you will always get the expected Run.
fastq-dl, is a spin-off of ena-dl (pre-2017), that has
been developed for usage with Bactopia. With this in
mind, EBI/NCBI and provide their own tools (enaBrowserTools
and SRA Toolkit) that offer more extensive access to
their databases.