[Bioperl-l] IUPAC support for DNA alignment

[aaron.j.mackey at gsk.com]

You could replicate what they do here with EST_GENOME (re-engineered to 
accept ambiguity codes):

  http://www.genome.org/cgi/content/short/17/2/212

But I think the answer is user-dependent -- some might want the "full 
score" (as in the above case), others might want the "(probabilistically) 
averaged score", etc.  So, let the scoring matrix be subclass-able (or 
mix-able), so that users can specify the exact desired behavior via a 
handful of predefined (and useful) behaviors.

-Aaron