question about sequence table

[Kuan Y. Chang]

Hi all,

I have a question about your sequence table.  I understand the need to
have sequence ontology, but I don't understand why GO decides to keep the
real sequences in the database.  In the May 2002 GO release, the sequence
data file is about 24 MB.  However, it seems that all of the 24 MB data
can be found thru their Swiss-Prot accession numbers.  Could someone
please explain why you keep the sequence data?  Thank you very much!

Kuan