Genius Drugs From Dumb Silicon

Can a giant pile of data beat human expertise in the design of miracle drugs? Daphne Koller may come up with a surprising answer.

- By Jillian D’Onfro

Can a giant pile of data beat human expertise in the design of miracle drugs? Daphne Koller may come up with a surprising answer.

NNot many scientists get solicited for photo ops, but for Daphne Koller it’s a regular occurrence. “It happens at pretty much any event that has tech people,” Koller says when asked about one recent snapshot. “It’s a little awkward. It’s not like I feel like this is something I deserve.”

Selfie requests are just one sign of Koller’s stardom, earned from more than 20 years bridging computer science, biology and education. She chalked up a string of accolades along the way: getting a master’s degree from Jerusalem’s Hebrew University at 18; becoming a Stanford University professor focused on machine learning at 26; winning, nearly a decade later, a MacArthur “genius grant” for research that combined artificial intelligen­ce and genomics; cofounding $1 billion (valuation) Coursera, an early platform to let people around the world take university classes for free.

The next act for this 51-year-old innovator: Insitro, a firm in South San Francisco that aims to find new drugs by sorting through masses of data. If it succeeds, it will have overturned how drugs get discovered.

Lab biologists typically focus on a few specific proteins as drug targets. If those fail, data scientists make suggestion­s for others to try. Insitro, on the other hand, wants to collect much more data before the biologists go off on their hunt. It will leverage advances in bioenginee­ring (such as Crispr gene editing) and in software that enables computers to see things that escape humans.

Koller describes her aha moment this way: “Machine learning is now doing amazing things if you give it enough data. We finally have the opportunit­y to create biological data at scale.”

Insitro’s computatio­nal experts and biologists work together to create lab experiment­s to produce massive custom data sets. Machine learning models then find patterns to suggest new tests and potential therapies. Robotics like automated pipetting machines reduce human error. With all this, Insitro can do “experiment­s in a matter of weeks instead of years,” Koller says.

AI plus biology, her background, was a “marriage made in heaven” for investors, she says. Within six months Koller raised $100 million from ARCH Ventures, Andreessen Horowitz, Foresite Capital, Alphabet’s venture fund GV and Third Rock, with Jeff Bezos and others joining later. In April, she landed a deal with Gilead Sciences that gives Insitro $15 million now with $1 billion to follow if it helps find a treatment for a deadly form of nonalcohol­ic fatty liver disease. The disease is expected to soon become the leading cause of liver transplant­s.

“There are very few individual­s who understand both sides of the beast,” says Mani Subramania­n, who heads liver disease clinical research at Gilead. “The biology as well as the deep learning.”

Insitro’s future payouts from Gilead hang on whether it can identify five proteins that could be targets for drugs and then whether targeting those proteins leads to approved therapies for the liver disease. The contingent payments, which include revenue sharing from successful drugs, helped Insitro earn a spot on Forbes’ inaugural AI 50 list of the most promising artificial intelligen­ce companies.

More than 20 other startups are chasing the dream of faster, cheaper drug discovery through AI. Among them are Notable Labs, with $55 million of venture capital, and Verge Genomics, with $36 million. Novartis has announced a five-year AI collaborat­ion with Microsoft, and Merck and GSK have startup partnershi­ps as well.

Artificial intelligen­ce does not make biology easy. “I don’t think the platform can be magic,” Koller says.

Before Insitro can reap rewards, a few hundred thousand lab tests need to happen. Koller has the energy. Bouncing around Insitro’s office—she gave away her desk chair to one of her 53 employees because she never used it—she moves from a room named Macrophage (a white blood cell) to one named Elastic Net (a data-modeling technique) to show off the latest lab equipment.

Big Pharma’s interest would seem to make Insitro a likely acquisitio­n target if it hits pay dirt. But Koller says she doesn’t want to see Insitro “swallowed into the maw” of a larger organizati­on. She wants it to make its own branded drugs.

The ultimate goal is that the people asking for photos ops will be healthier thanks to Insitro. Koller says she hopes they come up to her and say, “Because of you, I have my life back.”

“DATA TRUMPS EVERYTHING.” —Josh Estelle, a lead engineer for Google Translate

 ??  ??

Newspapers in English

Newspapers from United States