directory name as well as the file name, rather than $file, which
doesn't.

Use -d instead of -e to test for directories.

genes

present, and reuse it if it is.

file on FTP instead of from an HTML file.  Change populate_metadata.sql
and InterproParser.pm accordingly.

Update schema version.

store it as the release.

NB:  Since the release is data source specific (not species specific),
if you're running the parsing step of the Xref code for more than one
species, only the build number for the last species will be stored as
the release of UniGene.  So, keep species separate.

concatenated when using globbing.

Call set_release() for all relevant source IDs.

Remove UNIQUE constraint on translation_stable_id to bring it into line with the other stable_id tables.