Artificial intelligence is making better drugs

Niven R Narain

Niven R Narain
AI Drug Developer

WIRED 2015 is our annual two-day celebration of the innovators, inventors, artists and entrepreneurs who are reinventing our world. For more from the event, head over to our WIRED 2015 hub.

Creating new drugs is expensive. On average, it takes $2.6bn (£1.7bn) and between ten and 15 years to get a drug from planning to pharmacist, according to some estimates. And that’s because current methods are so untargeted — for every 100 drugs that reach first stage clinical trials, only one goes on to become an actual treatment.

“In any other industry in the world, we would all be fired,” saidNiven Narain, co-founder, president and chief technology officer at Berg, a company that’s pairing artificial intelligence with medical research to create more precise treatments.

Vincent Whiteman WIRED

“As an industry we have to do better but we have to go back to biology,” Narain explained. Berg’s approach is to create ‘maps’ of patients, creating huge quantities of data from large sample sizes, and mine that data for practical ways to make drugs cheaper, and quicker.

“We’re creating maps on every single patient. We’re taking all the complex biological information and transforming it into something that’s representative of the patient in the form of maps,” he told the audience at WIRED2015.

These maps take into account biological, clinical and real world information, building a narrative around each patient.

View gallery

“Precision medicine is going to be the future,” Narain argued. “We [need to] move away from this one size fits all [approach], because 90 percent of people with pancreatic cancer fail the first line of therapy.”

With the help of artificial intelligence, Berg and his team have developed a process to reactivate mitochondria in dying cells, stopping them from becoming cancerous. The drug, BPM31510, restarts the metabolisation of glucose and allows the body to harmlessly pass the problem cells. This drug, and the speed of its development, were only possible because of artificial intelligence and data analysis, Narain argued.

And the field is progressing quickly; Narain is confident that doctors will be prescribing drugs developed from similar systems in the next couple of years.

 “Artificial intelligence and the next generation of biology will allow us to participate in our own health,” he said.