[Bioperl-l] (no subject)

[Chris Fields]

On Jul 22, 2008, at 9:36 AM, Dhr. R. de Jonge wrote:

> Hi all,
> Thanks for the very quick responses. I've been trying to fix it today,
> trying both the bl2seq method and the direct approach. Unfortunately,
> I'm unable to fix either of the two.
> At the moment I'm trying to get the direct approach working, however,
> I'm not able to obtain the hit_start and hit_end numbers of the
> domain. It seems that these methods, which are not implemented in the
> Bio::Search::HSP::HMMERHSP module (which I used to obtain the domain

I believe the query is your sequence and the 'hits' are the HMMs (been  
a while since I used HMMER though).  In this case you don't want the  
hit start/end, you want $hsp->start('query')/$hsp->end('query') (or  
$hsp->query->start/$hsp->query->end, which should be the same thing).

> sequence). I've screened through the appropriate modules, with no
> luck. Is there a method to obtain these hit_start/hit_end figures by
> parsing? Or is there a better method?
> Many thanks again!
> Cheers
> Ronnie

See the HOWTO for SearchIO for more methods.

http://www.bioperl.org/wiki/HOWTO:SearchIO

Some hit/HSP methods for HMMER aren't implemented, IIRC.  For an  
alternative faster approach you can try Sendu's hmmer_pull  
implementation.

chris