UnivAln 1.004 Beta

[Georg Fuellen]

Steve wrote,
> > Arbitrary numbering schemes are now supported, you can write things like 
> > 
> >   $aln_copy->seqs({ids=>'Rattus Mus Junk'},{ids=>'1 1a Whatever 3 5 50'})
> > 
> 
> Um, I suggest that we leave the arbitrary numbering OUT of the interface
> for now, as it could potentially conflict with whatever is decided for
> Bio::Seq.  (Leave the code in, but don't advertise it)  There are a lot of

Well, no problem if the Bio::Seq beta is finished during the next few
weeks; I'll mark the relevant parts of the POD with an ``alpha!'' warning.

> questions which need to be very seriously thought about, such as what sort
> of assumptions should be made for efficiency and usability.  Just for
> example, should lists of positions be given as space-delimited strings,
> comma-delimited strings, arrays of strings, or something more complicated?

Note that if you accept a reference to a hash as the parameter,
the key makes it possible to support a lot of possiblities in parallel,
like ``{i=>'...'}'' for the final solution (whatever it is), or even
``{descs=>'...'}'' if you wanna support retrieval of sequences for which
the description matches a certain pattern. For ids which have no whitespace,
space-delimited strings look optimal. 

> > **** PLEASE, DO SEND ME FEEDBACK !!! ****
> > **** PLEASE, PROOF my POD and code !!!:-) ****
> 
> I'm just about to go on a 2-week lecture tour, so I doubt I can give
> comments before early April.  Maybe I can make some quick comments this
> weekend, but as I need to finish my slides and set off a paper -- don't
> bet on it.

Please give it a try :)

I'm really eager to not have another month-long (year-long..) delay,
so let's pls reserve time for Bio::Seq in April !

best wishes,
georg

> Steve
>