A key innovation of the new study is that the researchers were also able to figure out what kinds of information the deep-learning model was using to make its antibiotic potency predictions. This knowledge could help researchers to design additional drugs that might work even better than the ones identified by the model.

That is awesome. I wonder if the techniques that they have used to expose the machine learning “black box” process can be applied to other models - from my understanding of it, that would be pretty big news in and of itself.