traceback (was [Biophp-dev] scripts)

[Serge Gregorio]

>Frankie: from what i've learned so far alignment 
>can be done by variuous algorithms, needleman-wunsch
>being the most basic one described on the site u >linked to in a recent mail.

>http://molvis.chem.indiana.edu/C687_S99/lecture2.html
 #alignments)

Ah now it's clear... okay, I'll try to see what
I can dig up on this topic. If I can't give you  
links/materials myself, I know someone who can. =)

Personally, I am fascinated with the analysis and algorithmic side of bioinformatics.  But I realize some parts need to be developed first (like the parsers) before I jump into that.  Talk about saving the best for last, huh?  =)

>i attached a very basic bit that returns a matrix as >a result of a needleman-wunsch algorithm which could >be passed to a traceback function doing the actual >alignment. Of course its gotta be oo in the end, but
>i did tell u im lazy ;)

Hahaha, well, there's a lazy guy in all of us.  Did I mention that I stayed off hardcore coding for a whole week last week?  

Cheers!

Serge 

Frankie wrote:

--------- Original Message ---------

DATE: Fri, 30 May 2003 16:19:50
From: apu@gmx.li
To: Serge Gregorio <biophp-dev@bioinformatics.org>
Cc: 

>Hola serge,
>
>
>SG> Your code seems to be working fine.  I'll
>SG> do more tests on it.  Btw, could you be more specific as to
>SG>  what you were looking for last time?  The traceback thing?
>SG> I might have some links for you...
>from what i've learned so far alignment can be done by variuous
>algorithms, needleman-wunsch being the most basic one described on the
>site u linked to in a recent mail.
>(http://molvis.chem.indiana.edu/C687_S99/lecture2.html#alignments)
>All of them seem to be related, just varying in the different cost
>models they use.
>The traceback step comes after filling the matrix with the different
>costs for transformations, starting in the lower right corner one
>traces the possible alignments back to the upper left corner, by
>going up, left, or diagonal to where the lowest cell values are and
>diagonal having priority over the others. It is perfectly escribed in
>this article which uses a matrix build out of the levensthein
>distance.
>http://bioinfo.weizmann.ac.il/courses/BCG/lectures/02_pairwise/2.4tools/03DP.html
>
>Until now im not quite sure if the traceback process is the same in
>all those different global alignment algorithms, i might not know ;)
>If yes, one could have one single traceback function for all the different
>algorithms.
>Thats why im looking for material regarding the different algorithms, the
>ones i found where needleman-wunsch, smith-waterman and some
>extensions of them, which seem to share the Traceback procedure.
>



Need a new email address that people can remember
Check out the new EudoraMail at
http://www.eudoramail.com