Hi,

> 1. The annotations now come from 3 different "groups" or identifiers:
> 
> MAGPIE, reputer, calypso. (Please send me a reference for all three
> and a short 1 paragraph description for all 3 "methods"; thanks)

I have to check the references for calypso. I'm not sure if there are
any (this is a program Elizabeth wrote). Elizabeth will be back on
Thursday...

> 2. What is the difference between CDS and exon annotations? Original we
> had asked for exon annotations and if known additional CDS annotations 
> for all coding exons. What shall I use for "coding region" predictions 
> from MAGPIE?

CDSes are confirmed exons with (hopefully) correct exon boundaries and
splice sites. These CDSes have perfect (100%) matches against
databases sequences.

'Exon' features only have GENSCAN predicted exons, where the exact
boundaries and splice sites could not be confirmed, as we were not
using splice site prediction programs and no perfect matches have been
found.

So, use CDSes as "coding region" predictions from MAGPIE and 'exon'
annotations as possible (not confirmed) exons that have significant
hits in the databases. CDSes obtained by concatenating exons might have
wrong splice sites and frameshifts in the CDS...

> 3. "gene" feature. You use this as a line to functional annotate the
> per gene grouped entries. It has nothing to do with coding or
> non-coding. Correct?

Every 'gene' feature has the annotated function, regardless of
being predicted by GENSCAN ('exon' feature) or by MAGPIE ('CDS'
feature'). Every 'gene' is coding for its annotated function.


Cheers,

  Alex