[Bioperl-l] Bio::Align::AlignI for EMBOSS:needle

[Ryan Golhar]

That's exactly what I'm doing now....just regex parsing the similarity
line...didn't know if it was built into bioperl and I was just missing
it...

-----Original Message-----
From: Jason Stajich [mailto:jason.stajich at duke.edu] 
Sent: Monday, August 15, 2005 12:04 PM
To: golharam at umdnj.edu
Cc: 'Bioperl List'
Subject: Re: [Bioperl-l] Bio::Align::AlignI for EMBOSS:needle


You could sort of figure it out by processing the match_line output  
and counting the number of ':", '.', and '*' items and dividing by  
the aln length.

If we did parse it - there isn't really anywhere to put that sort of  
field right now in SimpleAlign.

I generally just have simple perl parser I run on needle/water output  
to get the percent similar/identical stats and if I need the  
alignment then parse it again with AlignIO;

Something like:
my %stats;
while(<$io>) {
  if(/^\#\s+(Identity|Similarity|Gaps):\s+(\d+)\/(\d+)\s+\(\s*(\d+\.\d 
+)\s*%\s*\)/ ) {
   $stats{$1} = [$2,$3,$4];
  }
}

$io->seek(0);
# process with AlignIO....



-jason

On Aug 15, 2005, at 11:38 AM, Ryan Golhar wrote:

> EMBOSS needle reports percent identity and percent similarity, however

> Bio::Align::AlignI has no method for obtain the percent similarity.
>
> My code is essentially:
>
> my $in = new Bio::AlignIO(-format => 'emboss', -fh => new 
> IO::String($output)); my $aln = $in->next_aln;
> $fepct = $aln->overall_percentage_identity;
>
> I tried the different percentage_identity methods to see if any of
> them
> work, but they don't give the similarity number.  Is there a way to  
> get
> the percent similarity through bioperl?
>
> Also, the description part of the document for 
> overall_percentage_identity has a type for the Title.
>
> Ryan
>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at portal.open-bio.org 
> http://portal.open-bio.org/mailman/listinfo/bioperl-l
>

--
Jason Stajich
Duke University
http://www.duke.edu/~jes12