With a small-molecule-ribosome complex crystal structure and historical data in hand, we use our sophisticated computational chemistry tools to rapidly design focused compound libraries. The approach converges on three areas: first, we use Analog to expand, search and score compounds in order to find substitute molecules for the bound antibiotic in the crystal structure with greater affinity for the ribosome. Here, we identify appropriate locations where we can decorate the molecule such that it can make additional interactions with the ribosome. As binding involves the bringing-together of the small-molecule and the ribosome, we evaluate simultaneously properties of the small-molecule outside of the ribosome in a model solution (think of the bloodstream, in which it is traveling) so as to understand what it will take to drive it to the binding site. This information, called descriptors, is grouped together, and a statistical model based on these descriptors is used to rank-order better-binding compounds. Second, we use Analog to optimize and score compounds in order to find molecules that have enhanced whole-cell activity across a wide panel of organisms. Here, we are looking not at the analog molecules in the ribosome but rather in a model solution (again, think bloodstream) in order to identify key features or properties that lead to activity across this panel. We prioritize compounds that not only bind well to the site of action in the ribosome but also that function to stop the ribosome from going about its usual business. Third, we use QikProp to compute properties of analog molecules that are well-accepted indicators of whether a molecule is likely to be an orally-available or an intravenously-administered drug. Compounds with the best profiles are advanced through development.