[ensembl-dev] transcript coordinates via REST?

[Reece Hart]

On Fri, Jul 31, 2015 at 2:38 AM, mag <mr6 at ebi.ac.uk> wrote:

> And a similar endpoint can provide you with transcript coordinates
> translated to genomic coordinates, although the reverse is not available
>
> http://rest.ensembl.org/map/cdna/ENST00000288602/100..300?content-type=application/json
>
> I am not sure if that covers your use case, so please don't hesitate to
> give us some feedback.
> The REST API is still in active development and we are always happy to get
> some feature requests that allow us to prioritise future endpoints.
>

Hi Magali-

My primary goal is to build a table of the correspondence between exons in
genomic and transcript coordinates so that we can map between g., c., and
p. sequence coordinates. The mapping is trivial in most cases, of course,
but indels and sequence errors are common enough that "most cases" isn't
good enough for diagnostic purposes. (For example, NEFL contains an insert
of a single G in a poly-G tract in GRCh37 that leads to a frameshift; the
refseq transcript is correct and doesn't have this pathology.)

NCBI started providing files with exon-level correspondence in April. These
are extremely useful. See
ftp://ftp.ncbi.nih.gov/refseq/H_sapiens/alignments/ for examples.

So, the request/suggestion is to provide a similar file or REST interface
for ENST transcripts.

While I'm at it, I would encourage Ensembl (everyone, actually) to use
sequence accessions (like NC_000001.10) rather than chromosome names.
Chromosome names are imprecise and unnecessarily create opportunities for
errors. (It also makes it hard to even consider a world in which multiple
assemblies are in one database schema.)

Thanks,
Reece
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20150802/51f874e0/attachment.html>