[EMBOSS] Help with stretcher

[michael watson (IAH-C)]

Hi Peter

The thing is, stretcher works fine for a lot of other examples (I'm
running a whole pipeline here) of the same problem; and I tried needle
but it complained the sequences were too long and that I should try
stretcher...

It would be good to remove this problem from stretcher given the size
limit on needle...

Many thanks
Mick

-----Original Message-----
From: Peter Rice [mailto:pmr at ebi.ac.uk] 
Sent: 02 November 2005 13:26
To: michael watson (IAH-C)
Cc: emboss at emboss.open-bio.org
Subject: Re: [EMBOSS] Help with stretcher

michael watson (IAH-C) wrote:

> Hi
> 
> I am using stretcher from emboss in a perl script.  I am trying to 
> globally align a relatively short sequence with quite a long one.
> Basically, the short one sits bang in the middle of the long one, with

> almost 100% identity.  However, no matter how much I play with the 
> gaplength and gappenalty options, stretcher insists that the the first

> base of the short sequence aligns with the first base of the long 
> sequence, followed by a HUGE gap in the alignment, followed by the 
> rest of the sequences, correctly aligned.  It's doing my nut in.

global alignemnts come ni two flavours - those that insist on aligning
the entire length of both sequences (like stretcher) and those that
allow gaps at the end without penalties (like needle). What you want
depends on the sequences you are aligning (i.e. what the ends really
mean)

You can simply use the needle program instead to get the results you
want.

Stretcher was originally part of the FASTA package. We can look at
adding an option for end weighting (and maybe making the default the
same as for needle).

Hope that helps,

Peter