The historical problem with mining the bacterial world has been that only about 1% of all bacterial species can be cultured in the laboratory. That was solved a couple of years ago with an ingenious low-tech approach called the Ichip. The people who solved it went on an ad-hoc, short fishing expedition and found a new viable antibiotic candidate the first time they tried. [1]
The future of antibiotics, like the future of a large chunk of drug development, is thus going to involve mining bacteria for the weapons they use on one another and the tools they use to eat things. We'll never run out of those, and we'll have more candidates than we know what to do with once people make a serious go of using the new methods of culturing bacteria. [2]
[1]: http://www.the-scientist.com/?articles.view/articleNo/41850/...