[Bioperl-l] Bringing Bioperl annotations in-line.

Mark Wilkinson mwilkinson@gene.pbi.nrc.ca
Fri, 06 Apr 2001 09:58:56 -0600


Ewan Birney wrote:

> It looks like we have a real, wholescale annotation rewrite in the
> off'ing. I think this is a Good Thing, as this is definitely an area
> where we've just thrown together objects that "work for one format".

I agree entirely - as I said in a previous message I (and Dave) are
constantly hitting the wall w.r.t. the usefulness of Tag/Value.  I'm
certainly willing to put in a fair bit of time to Get This Done Right, as it
will save me mountains of time later on!


> GO and Swissprot share a common idea about evidence tagging. We should
> probably just slip-stream their work here.

absolutely!  And this kind of thing can be easily abstracted... Chris earlier
pointed out that GO evidence tagging is very much like plain English:  <name>
said that <feature> was a <type> because they did a <exp_type> experiment:
<ref><xref>.  There is nothing particularly GO-specific about this...  I
suspect that most if not all evidence can be described in a minimal way with
just these terms.  Extending this will require that the Annotation object
know which ontology it is using, and the methods that this ontology allows.
As far as I am aware, Dave "Damian" Block has already written several modules
for Workbench which autoload/create methods on the fly - I suspect that he
will have a solution in his head for this problem already...


> Also play well with GFF.

<headache>GFF tag/values are not at all well suited to this kind of
hierarchical annotation, unless the underlying database is already
object-oriented and knows the meaning of the tag's.  This is all well and
good if you have bought into Acedb (which I love), but I can't imagine how
this is going to work in general...  unless you want to make a tag of
'Annotation' followed by a long free-text XML string</headache>

Did you have a solution in mind when you made that comment?  If so, please
share!  :-)


> Ok. So --- what I would like to suggest is that Mark, Hilmar, Myself,
> Chris and anyone else who wants to play come up with proposals (or modify
> proposals of each other) of real code

is it time to move this discussion off of the general list until we have
something more concrete to put forward for judgement/critique by the masses?


>    Bio::AnnotationI
>      defines ->get_ReallyInterestingStuff()

if only.... :-)


> Sounds fair?

sounds good to me.

M


--
---
Dr. Mark Wilkinson
Bioinformatics Group
National Research Council of Canada
Plant Biotechnology Institute
110 Gymnasium Place
Saskatoon, SK
Canada