[ensembl-dev] Retrieval of aligned transcripts using API

[Gregory Jordan]

You don't need to run an alignment for this task, as the relative position
of each transcript is well-defined by its genomic coordinates. What you
could do is extract each transcript in genomic space:

1) Collect the DNA sequence covering the entire gene
2) Replace any bases which are NOT exons in the current transcript with gap
characters

This will get you 4 strings of the same length but different patterns of
gaps and characters. Finally, treating these strings as your alignment, just
remove any columns which have a gap in all of the sequences. I think that
would yield the result you're looking for.

greg

On Tue, Feb 22, 2011 at 4:52 PM, Tjaart de Beer <tjaart at ebi.ac.uk> wrote:

> Hi all
>
> I would like to know if it is possible to get an aligned set of
> transcripts for a specific gene from one species from Ensembl?
>
> E.g. When looking at
>
> http://www.ensembl.org/Homo_sapiens/Gene/Splice?g=ENSG00000150540;r=2:138721590-138773930
>
> we see there are 8 transcripts of which 4 potentially codes for a protein.
> Transcripts ENST0000280097, ENST00000410115, ENST00000280096 will all
> align without any problem. The problem comes in when adding
> ENST00000329366. This transcript contains an exon that none of the other
> transcripts do. When using the sequences of the 4 transcripts in a normal
> alignment, it will try to align that exon with other pieces and not
> introduce a gap in the other sequences to accommodate the exon.
>
> Ideally the results should look like this:
>
> >ENST1
> AAC---CGAGGTTTT
> >ENST2
> AAC---CGAGGT---
> >ENST3
> AACCGGCGAGGTTTT
> >ENST4
> AAC---CGAGGTT--
>
> Using the API, is it possible to get this type of aligned transcripts back?
>
> Any other suggestions on how to do this are also welcome!
>
> Thanks!
> Tjaart
>
>
>
>
> _______________________________________________
> Dev mailing list
> Dev at ensembl.org
> http://lists.ensembl.org/mailman/listinfo/dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20110224/7e09da55/attachment.html>