[Dynamite] logging on from kinko's

[Ewan Birney]

;) I love kinkos. Ian - great work on the pages. 

Now for the bitching...

I really don't like the fact that SeqModule is dependent on
ScoreModule. Iwould like to be able to reuse SeqModule in
the complete absence of anything else. I realise that the problem
is ambiguous bases. Could we move the ambiguous base Score stuff
out somehow - eg - move these functions into the ScoreModule, and leave
alphabet with some representations about 'Y' -> 'C,T' or something
similar?

I think having SeqModule as a standalone module is a "good thing"
and should be tried for alomost at all costs...

Is SeqLikelihoodRatio a good name for that interface? It is sort of
an algorithm "bag" isn't it for one particular model. It is kind of
a particular dynamite.c file....

What about ModelMethods (?)


Not brilliant...


We haven't really cracked database searching yet either, but lets' walk
before we run.

I am still planning to flesh out the sequence stuff soon (un;less guy has
an overwhealming desire to). That metasequence and then move on to the
simple data structures in SingleModel. 

If anyone wants to join in on the implementation party speak up ;)


-----------------------------------------------------------------
Ewan Birney. Mobile: +44 (0)7970 151230
<birney@ebi.ac.uk>
-----------------------------------------------------------------