So just to clarify, is this what you mean?

/das/source/features?segment=P99999     [returns all positional and all 
nonpositional]
/das/source/features?segment=P99999:1,150     [returns only positional in the 
query range]
/das/source/features?segment=P99999;type=somenonpositionaltype     [returns any 
feature of the given type, which might be nonpositional]

This is different to what the spec says, but is close to what UniProt does. It 
does the above, except it also allows this:

/das/source/features?segment=P99999:0,0     [returns only nonpositional]

I think you're saying we should remove the 0,0 capability from MyDas/uniprot, 
and change the spec wording to reflect that nonpositional features will not be 
returned if the client specifies a start/end position. Right? :) Sorry for 
being dumb, am just a bit confused about what you're saying is illogical, etc.

On 2 Aug 2010, at 16:17, Jonathan Warren wrote:

> Currently having read all points and thought about  it properly I agree with 
> Eugenes summary of what the behavior should be.
> I was thinking that types filtering is not implemented and understood well by 
> many casual users of DAS. But as the types command and filtering is something 
> we want to encourage using the non-positional type should encourage this and 
> as long as it's documented well and splattered everywhere I think it should 
> be fine. But using the logic below as far as I can see this means you have to 
> do 2 requests if you want features and non-positional annotations for a range 
> /das/source/features?segment=P99999:1,150 and   
> /das/source/features?segment=P99999;type=annotation
> 
> But presumably doing 
> /das/source/features?segment=P99999:1,150;type=gene;type=annotation would 
> actually return genes if in region but no non-positional results which isn't 
> really logical on its own?
> 
> 
> On 2 Aug 2010, at 13:33, Andy Jenkinson wrote:
> 
>> On 2 Aug 2010, at 10:45, Jonathan Warren wrote:
>> 
>>> I'm not sure I agree with returning all annotations for every tiny part of 
>>> a sequence requested.
>>> If you consider DAS to be used for visual display mainly  - then it seems a 
>>> bit ridiculous to return all publications related to a segment (take a case 
>>> where you have many publications related to a protein sequence). If 
>>> publications aren't asked for i.e. non-positional annotations then I don't 
>>> think they should be given. So I guess I'm agreeing with Jim here.
>>> 
>>> Given the history of DAS I would propose a non-positional parameter as 
>>> apposed to "positional".
>>> 
>>> I think we have to remember that the 1.6 spec is supposed to mainly be a 
>>> consolidation of the way DAS is being used and DAS is supposed to be simple 
>>> (or not overly technical and difficult to pick up anyway). However- 
>>> obviously we don't want to propagate practices that really don't make 
>>> sense. The 0,0 thing is a hack like we had hacks for ontologies which now 
>>> for 1.6 we have cvIds (I think are a big improvement). So we need something 
>>> that is simple and obvious for a big all encompassing thing like positional 
>>> vs non-positional.
>> 
>> I'm not quite sure what you're suggesting we do in 1.6?
>> 
>>> 
>>> 
>>> 
>>> On 2 Aug 2010, at 09:50, Leyla Garcia wrote:
>>> 
>>>> *Hi list,
>>>> 
>>>> *>No magic numbers.
>>>> 
>>>> *According to discussions on this matter, I will change MyDas behaviour so 
>>>> 0 will be no
>>>> "magic number" any more,
>>>> 
>>>> *>Types can be used for filtering, and actually you get more fine-grained 
>>>> control than simply positional or non-positional. (I use this technique 
>>>> now in DASher.) *
>>>>> In my opinion, the current spec as written is correct. That is, 
>>>>> non-positional features don't just apply to the whole sequence, they 
>>>>> apply to any part of the sequence.
>>>>> As an example, consider a journal reference --- a particular protein was 
>>>>> isolated by a lab, they wrote a paper about it, and deposited the protein 
>>>>> sequence in a database. If you look at a subsequence of the protein 
>>>>> sequence, that subsequence still derives from the paper, right? So 
>>>>> therefore the feature containing that journal reference should still be 
>>>>> attached to the subsequence.
>>>>> On that basis, I think the uniprot server is technically doing it wrong 
>>>>> and should be changed, although I have to say that in practice it hasn't 
>>>>> been an issue for me.
>>>> 
>>>> *and non-positional features will be always returned.
>>>> Since UniProt is built upon MyDas, its behaviour will change as well.
>>>> 
>>>> Thanks,
>>>> 
>>>> Leyla
>>>> *
>>>> 
>>>>> 
>>>>> Dave
>>>> 
>>>> _______________________________________________
>>>> DAS mailing list
>>>> [email protected]
>>>> http://lists.open-bio.org/mailman/listinfo/das
>>> 
>>> Jonathan Warren
>>> Senior Developer and DAS coordinator
>>> blog: http://biodasman.wordpress.com/
>>> [email protected]
>>> Ext: 2314
>>> Telephone: 01223 492314
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> -- 
>>> The Wellcome Trust Sanger Institute is operated by Genome ResearchLimited, 
>>> a charity registered in England with number 1021457 and acompany registered 
>>> in England with number 2742969, whose registeredoffice is 215 Euston Road, 
>>> London, NW1 2BE._______________________________________________
>>> DAS mailing list
>>> [email protected]
>>> http://lists.open-bio.org/mailman/listinfo/das
>> 
> 
> Jonathan Warren
> Senior Developer and DAS coordinator
> blog: http://biodasman.wordpress.com/
> [email protected]
> Ext: 2314
> Telephone: 01223 492314
> 
> 
> 
> 
> 
> 
> 
> 
> 
> -- 
> The Wellcome Trust Sanger Institute is operated by Genome ResearchLimited, a 
> charity registered in England with number 1021457 and acompany registered in 
> England with number 2742969, whose registeredoffice is 215 Euston Road, 
> London, NW1 2BE.


_______________________________________________
DAS mailing list
[email protected]
http://lists.open-bio.org/mailman/listinfo/das

Reply via email to