But for the academics their prestige rests on publishing that mechanism of action. Determining x drug works via inhibiting such-and-such transferase via pi-pi

But for the academics their prestige rests on publishing that mechanism of action. Determining x drug works via inhibiting such-and-such transferase via pi-pi interactions (for example) is how careers are made, and in cases where the learning is important enough to shutter away, typically the academic will license it out or go into business for themselves -- which is just fine because that means we get a drug out of it, if there was one to be had. The problem is that the academic side isn't generally focused on things that matter about getting a drug to market(e.g. what happens when you swallow it? Does inhibiting such-and-such transferase actually do anything meaningful? What does it metabolize into? etc.) Lots more glitz and glamor (and lots cheaper) to make a compound inhibit something or kill cancer in a petri dish. Its vaguely as if say computer science academics were studying code golf.