University of Toronto startup Deep Genomics is using artificial intelligence to find better drugs for genetic disorders and to get them into the clinic faster.

Ever since scientists fully sequenced the human genome in 2003, they have been trying to figure out how to use the information to save lives. However, each person’s DNA contains six billion letters and there are tens of thousands of genetic disorders caused by millions of mutations, so pinpointing the cause of a genetic disease using traditional scientific methods is next to impossible—there is just too much information to sort through.

That realization was the key to the founding of Deep Genomics, a company launched at U of T that’s using artificial intelligence to pinpoint the genetic causes of disease and to rapidly discover and develop new therapies.

Artificial intelligence opens new possibilities in drug discovery for genetic disorders

“For the first time in history, the amount of biomedical data has exceeded the ability of humans to understand it and to act on it,” says the company’s founder and CEO, University of Toronto Professor Brendan Frey. “The way forward is to use artificial intelligence, which is very good at processing huge amounts of data, finding patterns and making accurate predictions.”

Deep Genomics is applying the techniques of deep learning, which is a type of artificial intelligence pioneered by Frey’s mentor, the renowned U of T professor Geoffrey Hinton. The idea is to use deep learning to mine the human genome for information on how diseases occur, and to design and test chemical compounds to treat them. Deep learning allows scientists to test billions of chemical compounds in a computer model and look for promising formulas that they could develop into life-saving drugs for people around the world.

Deep Genomics is widely considered a leader in its field. And the company that is scaling up drug discovery and development is growing by leaps and bounds itself. In early 2020, SpaceX investors backed Deep Genomics with $40 million in second-round investment financing.

Later in the year, Deep Genomics signed a deal with BioMarin Pharmaceutical to identify potential drugs for four rare genetic diseases. Also, their first clinical trial program is set to launch in 2021.

As well as growing their portfolio and capabilities, the company also recently expanded their leadership team, adding a new chief business officer and a chief medical officer.

“Amanda Kay, Ferdinand Massari (and all the leaders really!) join us at a critical inflection point for Deep Genomics, says Frey. “The evolution of our AI Workbench has allowed us to rapidly identify genetically defined targets and design compounds with high therapeutic potential. We are now advancing a broad pipeline of pre-clinical programs and plan a massive expansion of our partnership strategy.”

U of T’s world-leading AI network helped launch Deep Genomics

Deep Genomics’ own DNA is inextricably tied to U of T, which is a world leader in AI. It started with Frey’s PhD work with Hinton in the 1990s, when they published one of the first papers on deep learning in Science magazine in 1995. It is also reflected in the company’s staff today.

“U of T has some of the best scientists and engineers in the world, and Deep Genomics is fortunate to have been able to regularly recruit people from this pool,” Frey says.

The start-up got important early support from U of T’s entrepreneurship network, including the Creative Destruction Lab, the Innovations & Partnerships Office and UTEST.

“The U of T network offers highly effective channels for entrepreneurs to access unparalleled mentors and funding opportunities,” says Frey. “Deep Genomics was able to use these channels to raise funding on good terms, and to recruit world-class advisors. Toronto—and U of T—is possibly the best place in the world to start an AI-biotech company right now.”

Discover more stories