[Bioperl-l] Re: Another update for my alignment module

[Yee Man Chan]

> > >
> >
> > Ok. I can clarify by adding it is for single CPU...
> 
> I would just prefer you not say that any current algorithm "is the
> fastest", because it's bound to be wrong.
> 

I think it is the fastest right now. There can be faster one in the future
but not much faster because this is at least a O(m*n) problem. This is
proved analytically by Aho, Hirschberg Ullman in "Bounds on the complexity
of the longest common subsequence problem.", Journals of the ACM
23(1):1-12, 1976

> 
> We're talking about your package now, not SSEARCH; what search phase does
> dpAlign have?  Your synopsis says that I can get a Bio::Align-compliant
> object out of $factory->pairwise_alignment($s1, $s2) ... will I get an
> alignment object if I initialize my factory with DPALIGN_LOCAL_GREEN ???
> 

As of now, you will get a SimpleAlign object if both sequences are DNA.
There is no support for protein for Phil Green right now. I can add this
support once I figure out how to extract Phil Green's code from ssearch.

> The blosum62.mat file in Bio/Ext/Align/ has a row for '*'.

Thanks!

Regards,
Yee Man

> 
> -Aaron
> 
> -- 
>  Aaron J Mackey
>  Pearson Laboratory
>  University of Virginia
>  (434) 924-2821
>  amackey at virginia.edu
> 
>