Can AI Drug Development Live Up to the Hype?
Big Pharma is pouring money into tech-devised treatments, but the verdict is out on whether they’ll actually work.
Trying to get medicine approved by the US Food and Drug Administration often feels like buying an expensive lottery ticket, with the uncertain prospect of a reward years away. It takes about a decade and more than $1 billion on average to take a drug through all the stages of clinical trials, and about 90% never get that far.
Part of the problem is the way researchers discover treatments. Until recently, scientists would come up with a hypothesis and run numerous tests to find a chemical capable of shutting down, say, a disease-causing protein. Then they’d run additional tests to see if they could make that drug stronger, safer and longer-lasting — a risky and time-consuming process. Now the pharmaceutical industry is hoping to make the whole program faster and more predictable with artificial intelligence.