by Michael Keller
An infectious disease researcher named Eve has made a surprising discovery—a compound named TNP-470 that stops tumors from growing also shows promise in fighting malaria.
It’s an important early-stage drug development find. Around 200 million people suffer from malaria around the world and half a million die from it every year. But this isn’t a story about TNP-470; it’s about Eve.
The scientist worked day and night in a University of Manchester lab in the UK to uncover this promising lead. To link the compound with a new disease target takes a lot of mind-numbing, repetitive tasks from the earliest stages of work. Either in a lab experiment or on a computer, compounds are run through initial tests to see if they might do something useful against a disease. All those that might show promise, which could be in the thousands or millions of experimental compounds, are then tested again in a mass screening. Some will appear to do something useful against the disease; others that must be weeded out will offer false positives. This subset is then retested to confirm the results.
Hurdles continue to be thrown at candidate compounds and the reject list expands until just one really promising molecule remains. This long process is part of the reason why it can take more than 10 years and $1 billion to develop a new drug.
But Eve didn’t care about all the work that goes into the basic research behind drug discovery. And she sped up the discovery curve as she worked, learning from early tests to do better in later ones. When the sun went down and other researchers went home for the evening, the scientist kept working. She didn’t take a break to get caffeinated, or even to turn the lights on. That’s because Eve is a robot, with arms to manipulate samples and advanced instruments to analyze them, that learns as it processes compounds through complex artificial intelligence.