[Biojava-dev] Java implementation of Smith-Waterman algorithm

[Ahmed Moustafa]

Hi Mark,

Which  object does hold the output of the alignment?

Thanks,
Ahmed

Schreiber, Mark wrote:

>Hi -
> 
>Because we have traditionally used HMMs there where no scoring matrices per se, just transition and emission probabilities.
> 
>A scoring matrix would be easy to acheive. I would suggest you start with an interface called ScoringMatrix that would have as a minimum methods like:
> 
>String getName(); which would return Blosum60 or similar depending on the implementation.
>int getScore(Symbol s, Symbol substitute) which returns the int score for the substitution.
> 
>You might want to also make a ScoringMatrixFactory that can generate appropriate instances of the ScoringMatrix from an XML file or files that could be included in the Distribution.
> 
>- Mark
> 
>
>	-----Original Message----- 
>	From: Ahmed Moustafa [mailto:ahmed at arbornet.org] 
>	Sent: Tue 14/10/2003 6:45 p.m. 
>	To: Schreiber, Mark 
>	Cc: biojava-dev at biojava.org 
>	Subject: RE: [Biojava-dev] Java implementation of Smith-Waterman algorithm
>	
>	
>
>	Hi,
>	
>	What are the Biojava's objects for the scoring matrices e.g. BLOSUMs and
>	PAMs?
>	
>	Thanks,
>	
>	Ahmed
>	
>	On Mon, 13 Oct 2003, Schreiber, Mark wrote:
>	
>	> Hi -
>	>
>	> Probably just convert to use BioJava Symbol and SymbolList objects. Probably everything else is OK.
>	>
>	> - Mark
>	>
>	>
>	> > -----Original Message-----
>	> > From: Ahmed Moustafa [mailto:ahmed at arbornet.org]
>	> > Sent: Sunday, 12 October 2003 7:06 p.m.
>	> > To: Schreiber, Mark
>	> > Cc: biojava-dev at biojava.org
>	> > Subject: RE: [Biojava-dev] Java implementation of
>	> > Smith-Waterman algorithm
>	> >
>	> >
>	> > Hi Mark,
>	> >
>	> > I believe the current API is reusable. Is it necessary to
>	> > convert the already existing implementation?
>	> >
>	> > Anyway, how can I convert my classes to biojava?
>	> >
>	> > Thanks!
>	> >
>	> > Ahmed
>	> >
>	> >
>	> > On Sun, 12 Oct 2003, Schreiber, Mark wrote:
>	> >
>	> > > Hi -
>	> > >
>	> > > We have traditionally done pairwise alignments using HMMs. However
>	> > > there have been numerous requests for an implementation of
>	> > > Smith-Waterman. If you want some help coverting your classes to
>	> > > biojava give us a yell on the list.
>	> > >
>	> > > - Mark
>	> > >
>	> > >
>	> > >   -----Original Message-----
>	> > >   From: Ahmed Moustafa [mailto:ahmed at arbornet.org]
>	> > >   Sent: Sun 12/10/2003 9:19 a.m.
>	> > >   To: biojava-dev at biojava.org
>	> > >   Cc:
>	> > >   Subject: [Biojava-dev] Java implementation of Smith-Waterman
>	> > > algorithm
>	> > >
>	> > >
>	> > >
>	> > >   Hello,
>	> > >
>	> > >   I am working on a Java package of implementations of
>	> > sequence alignment
>	> > >   algorithms. I have released an implementation of Smith-Waterman
>	> > >   algorithm with Gotoh's improvement. The time complexity
>	> > is O(n2) and the
>	> > >   space complexity is O(m * n + n) .
>	> > >
>	> > >   The package name is JAligner and it is hosted at sourceforge.net
>	> > >   <http://jaligner.sourceforge.net>. There is a front-end
>	> > demo using Swing
>	> > >   and Java Web Start.
>	> > >
>	> > >   Could JAligner be incorporated into the BioJava project?
>	> > >
>	> > >   Best Regards,
>	> > >
>	> > >   Ahmed
>  
>