Analysis of predicted protein sequences.

[dan]

Hello all,

I have recently been fortunate enough to clone a novel gene, I have 2.7kb
of coding sequence and a predicted translation.

I am looking for analysis programs that will help me determine possible
functions for the protein.

However, when I say it is a novel gene, it is a novel gene - and I can't
find anything like it in the protein databases.  I have used most suites
for motif recognition / homology comparison / secondary structure
prediction, but all I can find is a potential NLS... 

If anyone has any suggestions for WWW / email servers that they are
particularly impressed with or use regularly and wouldn't mind a little
correspondance about the output of such systems I would be very grateful.

As an aside the N-terminus is serine/threonine rich, with three stretches
of 7+ serines in a row.  There is also a single stretch of histidines
(again 7+) which as a motif seems to be present in some classes of 
homeobox genes, but *outside* the box region.  If anyone could shed some
light on potential roles / interactions of these regions I would also like
to know.

Thanks in advance,

Daniel Swan
Graduate student (47 days left and counting)
MRC Clinical Sciences Centre,
ICSM Hammersmith Hospital,
London.
+44 (0) 181 383 8298