2
$\begingroup$

I would like to assign a URL/URI to accession runs (i.e. SRR IDs) from the NCBI short read archive.

In program listings, I used to list them as ftp:// URLs, because they were available on their servers as such, e.g. SRR3418242:

ftp://ftp-trace.ncbi.nlm.nih.gov/sra/sra-instant/reads/ByRun/sra/SRR/SRR341/SRR3418242/SRR3418242.sra

But this approach is deprecated, plus, not all runs are available this way anymore. The now recommended approach is to use sra-tools to download the data. This tool takes the SRR Ids directly, i.e. prefetch "SRR3418242", and resolves them to one of many possible locations (home directory, http, etc).

Is there an official URI scheme that can be used to reference accession runs from the short-read-archive ? I would like to disambiguate them from other types of files and URLs (S3 keys, http, file). For file manifests, spreadsheets, and publications, is there something similar to srr://SRR3418242?

$\endgroup$
2
  • 3
    $\begingroup$ No, there isn't something like srr:// and there won't be. ENA via aspera is almost always the preferred way to download data as long as ENA has it. $\endgroup$
    – user172818
    Mar 21 '19 at 22:21
  • 1
    $\begingroup$ Thanks. My concern is the namespace. I'd like to know a nomenclature which doesn't prescribe a particular software or particular read archive. As long as run numbers and sequence data agree on both archives... $\endgroup$
    – init_js
    Mar 22 '19 at 19:40
1
$\begingroup$

The nomenclature is no longer being followed as NCBI is moving all the files to S3/Google cloud. pysradb allows you to fetch the metadata and get the URLs for each individual SRR. See examples in this notebook.

You can also download an entire project by doing:

pysradb download -p <SRP>
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.