[Biopython-dev] DAS client in biopython

[Peter]

On Thu, Jul 22, 2010 at 3:13 PM, Andrea Pierleoni
<andrea at biocomp.unibo.it> wrote:
>
>> The heirachy seems unnecessarily nested, why not move the
>> code in Bio/DAS/DASpy.py into Bio/DAS/__init__.py? Or
>> even into Bio/DAS.py instead? Then that import becomes:
>> from Bio.DAS import DASpy, which also avoids the ambiguity
>> of DASpy for a module and a class. Are you expecting to have
>> other files under Bio/DAS?
>>
>
> hierarchy is now simplified to a single file DAS.py under Bio.
>
>> Also the name DASpy confuses me, maybe the class
>> should be something about DAS Servers?
>>
>
> I renamed the DASpy class to DASregistry so the main call now
> is:
>
> from Bio.DAS import DASregistry
>
> das = DASregistry()
>
> simplier...
>

That appears to make sense :)

>> Would it be right to regard the class DASSeq as a subclass
>> of SeqRecord? It looks like a minimally annotated sequence.
>> See also the DBSeqRecord in BioSQL.
>>
>
> I've been thinking about it and, actually, the DASSeq class
> corresponds exactly to information and methods available in
> the DAS sequence method, so I'd leave it this way.

So what DAS calls a sequence is closer to Biopython's SeqRecord
than Biopython Seq object? Hmm - that could cause confusion,
whatever you call your class.

> Most of the time this class shouden0t be accessed. and a clean
> SeqRecord object can be obtained using the "fetch_to_seqrec"
> method in DASregistry.

I'll take another look at your code later.

Peter