[ssml] Re: Sequence defined domains?

[Dan Bolser]

<quote who="Robert Finn">
>
> Hi Dan,
>
> HMMs should model the situtation that you describe.  We even force some of our
> HMMs to represent discontinous domains by making it ignore the nested domain.
>
> Cheers,
>
> Rob
>
> On Thu, 27 Nov 2003, Dan Bolser wrote:
>
>> Hi,
>>
>> When defining domains based on sequence searching, do you find the 'dom-in-dom'
>> type situations which are seen with structural domains?
>>
>> For example at the chain level
>>
>> --/---A-part1---/----------B---------/----A-part2----/---
>>
>>
>> which forms two globular domains at the structural level.
>>
>> I guess HMM can detect this, in addition to the situation
>>
>> --/----------A-alone-----------/---
>>
>>
>> Is it a part of Pfam or any other sequence based domain definition?
>>
>> Cheers,
>> Dan.
>>
>>
>
> ====================================================================
>
> Robert Finn
> The Wellcome Sanger Institute
> Wellcome Trust Genome Campus
> Hinxton
> Cambs
> CB10 1SA
> Tel:+44 (0)1223 495330
> ====================================================================