Re: [boost] Bioinformatics algorithms in boost?

From: Robert Goldwein (robert.goldwein_at_[hidden])
Date: 2006-12-05 13:19:29

Dear Leo,

you expressed that perfectly, that's _exactly_ my goal. Also,
bioinformatics algorithms are based on many general algorithms, and
advances even in more specific areas (such as some good solution for
global alignment problem) would be good for other areas.

Thanks for your nice reply, it's good to know I chose the right way.

Robert

-----Original Message-----
From: boost-bounces_at_[hidden]
[mailto:boost-bounces_at_[hidden]] On Behalf Of Leo Goodstadt
Sent: Tuesday, December 05, 2006 15:53
To: boost_at_[hidden]
Subject: Re: [boost] Bioinformatics algorithms in boost?

 > > Robert Goldwein wrote:
 > > > Would be there - in some near future - any interest in such
 > > library?
 > However, I also think it may be a bit too domain-specific to belong
to
 > Boost.

Dear Robert and Johan,

I too think that it would be great to have a bioinformatics library. 
Parts of it may ending up being quite domain specific but there are many

algorithms which are also of general interest. Hamming distance & 
Longest common subsequence are obvious examples. HMMs and dynamic 
programming also have much general utility. Slightly more esoteric 
things (?) like MCMC, EM, various sorts of ML are becoming more common 
place as everyone tries to model everything (e.g. in finance).

Part of the attraction of trying to do it in c++ is to see how far 
absolute performance can be maintained without compromising flexibility 
and a good design.

At the moment, there is little self-contained well designed good 
performing code freely available for many of these routines in any 
language. Having it in c++ would be a major boon.

Leo Goodstadt
_______________________________________________
Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost