The download/domains/ directory contains conserved protein domain/motif
information.

Protein domains were predicted using IprScan software from the InterPro
database. For more information about the InterPro database and 
for help interpreting IprScan output, see:

ftp://ftp.ebi.ac.uk/pub/software/unix/iprscan/README.html

Individual species-specific subdirectories contain the downloadabble files, and README documents with more specfic information about the contents.