Through horizontal gene transfer, they can exchange gene fragments with other bacteria until the right combination is found to prevent them from dying. In the case of viruses, they can change, through a built-in mechanism, their outer protein coat. And drug discovery is entirely trial and error. Hundreds of small containers with a sample of the illness causing material are exposed to many different chemical combinations. Then they are scanned for results. That’s why drug development is so expensive. So, X number of sample chemical combinations did nothing to harm the sample, a number did a little harm, and a number did a lot of harm or killed the sample. Now, the problem becomes, how will it affect a living animal that has the disease? They don’t know. If the trials on animals for X number of chemicals kill the animal and the disease, that group is out. If the disease is harmed but the animal experiences too many harmful side effects, then that group is out. Only the remaining chemicals that produced the least number of harmful side effects are candidates. And even after a drug is approved and it works, on TV, they will tell you the list of possible side effects which, in all cases, sound worse than the cure! Would I ask my doctor about that drug? No. Of course not. But there it is, if you’re willing to take a chance on not getting the worst side effects. There is no guide - no cookbook - no predictive manual.