Bernd Web | 18 Sep 19:33

SimpleAlign/AlignIO

Hi,

Possibly I overlooked this, but is it not possible to supply an
alphabet to SimpleAlign?
When SimpleAlign starts making sequence objects _guess_alphabet is
executed and in case the sequence (slice) only contains gaps a warning
is raised as the alphabet cannot be guessed.
As all sequences in the alignment (generally) would be of the same
type, it could be good to set the alpabet for the alignment; this also
would prevent the executing of _guess_alphabet for all sequences. Is
this possible?

Regards,
Bernd
Chris Fields | 18 Sep 19:56
Gravatar

Re: SimpleAlign/AlignIO

Bernd,

I'm not sure it's a good idea to constrain the entire alignment (and  
all sequences contained within) to a single alphabet.  For instance,  
what if I wanted to align a translated (protein) sequence, possibly as  
3- or 6-frame, to a DNA sequence of interest?  Or if I wanted to align  
a Meta sequence w/o an alphabet (i.e. a structural consensus sequence)  
with a DNA/RNA/protein sequence?  The former example is fairly common  
and may not be entirely possible with SimpleAlign (I haven't tried it  
to tell the truth), but the latter is what I do with the Stockholm  
parser.

chris

On Sep 18, 2008, at 12:34 PM, Bernd Web wrote:

> Hi,
>
> Possibly I overlooked this, but is it not possible to supply an
> alphabet to SimpleAlign?
> When SimpleAlign starts making sequence objects _guess_alphabet is
> executed and in case the sequence (slice) only contains gaps a warning
> is raised as the alphabet cannot be guessed.
> As all sequences in the alignment (generally) would be of the same
> type, it could be good to set the alpabet for the alignment; this also
> would prevent the executing of _guess_alphabet for all sequences. Is
> this possible?
>
>
> Regards,
(Continue reading)


Gmane