[Bioperl-l] First commit of Bio::Structure objects

Elia Stupka elia@fugu-sg.org
Wed, 21 Nov 2001 02:32:33 +0800 (SGT)


> I still haven't found a decent way to look at an 'id' in this case. A
> residue for example only makes sense in the whole context. I.e. "it is
> Val-42 from chain B from the third model from PDB entry 1XYZ". This
> would leave us with something like "1XYZ.3.B.VAL-42".

> I'm not sure the Ensembl (database ?)-centric way of looking at this is
> the right thing. My first plan was to use references as unique ID's.

I realised this after I wrote it, I was thinking of "assigned ids" or
database internal ids, becomes even more difficult if it is not database
related, although I must say with this kind of structure it seems to me
much lighter to pull out the graph structure based on ids, and only fill
the objects when you need them, otherwise they risk to become very heavy
objects I think...

> Can you elaborate a bit more on your proposal as I can't quite grasp it

Sorry, it was nothing special, all I was saying was that if we were
talking about ways to "hack" the problem, and if one had some kind of id,
then one could just fill each object with a get/set of the id for the
parent needed, thus allowing easy retrieval of the parent without circular
references.

Elia

-- 
******************************
* http://www.ebi.ac.uk/~elia *
* tel:    +65 874 1467       *
* mobile: +65 90307613       *
* fax:    +65 777 0402       *
******************************