Greetings all,
Though there are a number of modules available at CPAN relating to PHP, PHP and PHP::Include::Vars to name a few, my question to you is why are you not able to extend the objects offered in BioPerl to work with DBI, GD and Archive::Zip? Since you are using BioPerl in the first place it sounds as though your project manager is not opposed to OpenSource software and you should be able to, in theory as Im only guessing at this point, download and use other free modules like those from CPAN... or am I missing something?
Looking at the specs for BioPerl 1.4 on BioLinux.org I see that GD is bundled inside and that DBI and DBD::MySQL are suggested, so the developers understood that people would need graphics and DB interaction. It seems like it would be easy to create your own objects that use BioPerl along with your own DB and Archiving subroutines to get the desired functionality that you want.
Basically what Im trying to say is "don't through the baby out with the bathwater", as the saying goes "There is more than one way to do it" and sometimes those ways are not the most obvious.
Beside Perl Objects are beautiful things once you start writing them.

Update!
Here is an article about how to run Perl through PHP for those interested.

-InjunJoel