I'm writing some scripts to download data. Specifically, I need FASTA versions of: * All the "finished" mouse proteins in refseq * All the "finished" human proteins in refseq * All the sequences in dbSNP Ideally, my script would produce updated versions of these datasets nightly or so. I would prefer to do this without spamming the NCBI servers (or my bandwidth providers) too much. I've messed around with the bioperl Bio::DB routines enough to get really confused by ENTREZ queries. I've also looked at the FASTA source available through FTP from NCBI, and that confused me more. How do smart people do this sort of thing these days? -Chris Dwan