The assays that I have in mind are disk diffusion assays and assays to determine minimum inhibitory concentration or related. Sometimes these assays are conducted in rich medium (Luria broth, Terrific Broth, or similar). It seems to me that this choice biases against finding inhibitors that target enzymes whose absence is conditionally lethal. If one inhibits an enzyme that produces a metabolite that can be obtained from the medium, then growth of the cells might be impaired little or not at all. If one is interested in developing an antimicrobial compound, perhaps using a medium that mimics human serum is a better choice, but I imagine that there are complications I doubt very much that I am the only person who has wondered about this possible problem. Are there any papers that touch upon this subject or address it in a systematic manner.