Feature inheritance with sequences

[Peter Rice]

David Martin wrote:
> 
> On Fri, 9 Feb 2001, Peter Rice wrote:
> 
> > In principle, yes. Features in EMBOSS follow the GFF model, so they are stored
> > as groups of start and end positions. Assuming that they are not one of the
> > more exotic kinds of start and end positions, it is easy to 'insert' one
> > feature table into another. Deletion is also possible, although ends of
> > features within the deleted region rules out making the process completely
> > reversible.
>
> But does it actually work in practice? If one creates a new sequence from
> a and b, do the features travel with it?

Depends on whether the sequences have features. At present, you can turn on
features for an input sequence, but most applications don't use it (wisely, as
I am rewriting it heavily).

> Will EMBOSS write out features as standard in EMBL/Genbank format files or
> are we stuck in having extra GFF files floating about..

They are stored internally as 'GFF with hints to EMBL feature locations' but
can be written out in any format - including formats that have not yet been
invented. Some of those formats could even be read back in again. Others will
be standard forms of application output. Yes, that does include
EMBL/GenBank/Swissprot entries with feature tables.
 
> It would be good to have a tutorial on features.

I will need to write one for Alan anyway once the updates are done.

-- 
------------------------------------------------
Peter Rice, LION Bioscience Ltd, Cambridge, UK
peter.rice at uk.lionbioscience.com +44 1223 224723