[ensembl-dev] versioned Ensembl identifiers in HGVS tags reported by VEP

[Reece Hart]

On Mon, Jan 16, 2012 at 3:04 AM, Kieron Taylor <ktaylor at ebi.ac.uk> wrote:

> It is true that versioned identifiers are not really intended for external
> use.


On Mon, Jan 16, 2012 at 3:12 AM, Will McLaren <wm2 at ebi.ac.uk> wrote:

> we are required by HGVS standards to provide a
> version number with any transcript or protein identifier that we list
> as part of HGVS notation.
>

Kieron, Will-

Thanks for the quick explanations. I was afraid I was missing something
about identifier versions and the API.

The language regarding versioned identifiers is *should*, not *must*. The
HGVS "nomenclature" was written with NCBI in mind, and this inclination has
become increasingly explicit over time. For example, the General
Recommendations (http://www.hgvs.org/mutnomen/recs.html) state "the DNA
reference sequence used should preferably be from the RefSeq database,
listing both database accession and version number (like NM_004006.2)".
Other areas suggest "accession.version" but don't recommend NCBI, which
gives the impression that one should conjure up a version from somewhere.

VEP is terrific. Thanks!

-Reece
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20120116/49994d7b/attachment.html>