[Bioperl-l] [Fwd: [Bioperl-guts-l] Notification: incoming/1023]

Heikki Lehvaslaiho heikki@ebi.ac.uk
Fri, 26 Oct 2001 10:35:39 +0100


I checked this one (see below) and I agree that Lize is right. (consensus_aa
is in fact internal method _consensus_aa. The public method is
consensus_string().)

If consensus threshold is set to 100% then the returned string should
logically have residue characters on positions where all sequences are
identical. At the moment the program places '?' characters in every
position. The documentation for the method is consistent with this
behaviour.

I am posting to the main list before fixing this because 

1.) fixing this changes the output from consensus_string()
2.) Ewan who wrote it might have a good reason for this?

	-Heikki


-------- Original Message --------
Subject: [Bioperl-guts-l] Notification: incoming/1023
Date: Thu, 25 Oct 2001 15:56:31 -0400
From: bioperl-bugs@bioperl.org
To: bioperl-guts-l@bioperl.org

Full_Name: lize 
Module: SimpleAlign.pm
Version: 0.7.1
PerlVer: 5.6.0
OS: linux
Submission from: (NULL) (61.135.52.3)


the consensus_aa function

if we set threshhold 100,
then the output string will be all ?

this is because the author assume 
the number of same letter at a point must be   bigger than the
threshhold*number_of_sequence/100, but omitted the equl situation.


_______________________________________________
Bioperl-guts-l mailing list
Bioperl-guts-l@bioperl.org
http://bioperl.org/mailman/listinfo/bioperl-guts-l