[Bioperl-l] Pairwise seq. comarison

Jason Stajich jason at cgt.duhs.duke.edu
Fri Apr 2 17:19:20 EST 2004


Barry - What do you mean vague....? We all love to document our code... =)


FYI If you thinking popgen type stuff so finding number of polymorphic
sites?  Most of this code is pretty new but in Bio::PopGen - the latest
stuff in Bio::PopGen::Utilties allow for generating
Bio::PopGen::Individual individuals from a MSA alignment.

But since you are looking at pairiwse you probably just want raw
differences for a pairwise sequences you can do this:

my $stats = Bio::Align::PairwiseStatistics->new();
Actually depends on what you want to do

my $differences = $stats->number_of_differences($aln);

Note this is case-sensitive and doesn't do any fancy ambiguity code stuff
or anything.


If you want to do more fancy DnaDistance calculations which attempt to
correct for various biases Bio::Align::DNAStatistics has most of that
stuff.

This stuff has been under fairly active development post 1.4 release so
what is in CVS may be more accurate and correct.

-jason
On Fri, 2 Apr 2004, Barry Moore wrote:

> Bio::Align::PairwiseStatistics has a method "number_of_differences".  I
> haven't used it, but I presume you'd use AlingIO to bring in you
> alingment, and then just call the "number_of_differences" directly on
> your Bio::Align::AlignI  object.  The docs look kind of vauge, so
> hopefully someone else can flesh out this suggestions, but if your just
> looking for where to start experimenting, that's one place to try.
>
> Barry
>
> henrik nilsson wrote:
>
> >Hi,
> >
> >For two aligned (and obviously overlapping) sequences, can I use Bioperl to
> >retrieve the number of differing sites in the alignment (for the overlapping
> >region only, excluding the flanking regions where only one of the two
> >sequences is present)?
> >
> >Thanks,
> >
> >RN
> >
> >
> >_______________________________________________
> >Bioperl-l mailing list
> >Bioperl-l at portal.open-bio.org
> >http://portal.open-bio.org/mailman/listinfo/bioperl-l
> >
> >
>
>

--
Jason Stajich
Duke University
jason at cgt.mc.duke.edu


More information about the Bioperl-l mailing list