how to use your own set of seqs as the database to search??

[being]

we would like to be able to have a data file containing a set of
sequences to use as a searchable database using FASTA or some other
search program in gcg.

how does one set up such a database in your own dir?  

i hope this is clear.  instead of searching the database normally
searched by the fasta or blast program we would like to specify a much
smaller database of our own creation.

thanks for any help you can offer folks.