Artificial Intelligence, BioPharma

AI-biotech Terray emerges with $60M and a solution for data problems in drug R&D

Terray Therapeutics uses tiny microarrays to test molecules against targets of interest, then applies artificial intelligence to build large chemical datasets. The startup’s Series A financing will support R&D initially in immunology; the first Terray molecule is expected to reach the clinic in 18 months.

 

Artificial intelligence-driven drug discovery can be like conducting an internet search. With each query, your search engine learns a bit more about you—information that shapes the next set of results, said Jacob Berlin, co-founder and CEO of Terray Therapeutics. The same thing happens when you scan Netflix for movies, or shop on Amazon.com. The algorithms may not be very good at predicting what you’re looking for at first, but with more searches and more time, they refine their recommendations.

“That opportunity is there for drug discovery but only if we have that data,” Berlin said.

There are plenty of companies that are developing artificial intelligence technologies for drug research, aiming to predict which molecules are the most promising for therapeutic applications. But Berlin contends that the fundamental problem with AI in drug discovery is not in the algorithms, it’s the data, or rather the lack of it.

Terray generates the data that fuel its drug research, and the company is gleaning better and better insights from its analysis to find small molecules that address tough-to-treat diseases. The Pasadena, California-based startup has been quietly developing its technology for the past three years. The company emerged from stealth Tuesday backed by $60 million in new funding and plans to reach the clinic in less than two years.

Terray stems from Berlin’s effort to overcome challenges he faced earlier in his career. He recalled his experiences doing small molecule research at Harvard, MIT, and Caltech. Berlin said the process was slow, and even when he was able to find success, the results left many problems unsolved. He wanted to figure out a way to discover small molecules, but more quickly and at scale.

The solution that would become Terray is the culmination of research that began during Berlin’s time running a lab at City of Hope. There, his research group placed molecules in microarrays, each one about the size of a nickel. Those microarrays were then used to analyze the molecules and choose the ones for further study.

In 2018, Berlin co-founded Terray with his brother, Eli Berlin, the company’s chief financial and operating officer. Similar to Jacob Berlin’s City of Hope research, Terray also uses microarrays, small chips with 32 million tiny wells, each one containing a molecule of interest. The company’s technology, called tNova, exposes those chips to a target of interest and measures the binding affinity of each of the molecules. Each chip can be read in less than five minutes, Eli Berlin said. In a single day, tNova can measure billions of data points. Just as the Netflix and Amazon suggestions improve with more data and more time, so too do the molecule predictions from Terray’s technology. The accumulated data become part of libraries of molecules, which the company’s technology builds in the span of weeks.

“AI alone is not a solution, AI is a tool that works with the data,” Jacob Berlin said. “When those two match, the opportunities are endless. If you have the right type of data and the right AI and machine learning tools, they work together. The data get better, the predictions get better.”

The Terray technology is not limited to any particular therapeutic area or target. The company’s initial focus is immunology, which is dogged by selectivity problems—molecules hit other places in addition to their intended targets, resulting in side effects. Jacob Berlin said that by measuring many molecules and building maps showing how they interact with targets, Terray is able pick selective molecules from the outset.

The targets of those molecules, and the diseases the company aims to treat, remain undisclosed for now. Jacob Berlin would only say that Terray is developing small molecules that go after some of the hardest and undrugged targets in immunology. The startup’s progress has attracted interest from some biotech and pharmaceutical companies that have signed on as partners, using the tNova to address their drug discovery challenges.

With the new capital, Jacob Berlin said Terray will be able to advance to the preclinical research that could support an FDA application to proceed with human testing. Terray’s first molecule could reach the clinic in 18 months. In addition to the internal drug research, Eli Berlin said the new capital will support the startup’s efforts to pursue more work with partners (current partners remain undisclosed). Terray currently employs 50. With the growing internal research and partnership work, Eli Berlin projects that headcount could double by the summer of 2023.

The Series A financing announced Tuesday was led by Madrona Venture Group. Terray had previously raised $20 million in an undisclosed seed financing, co-led by Digitalis Ventures and Two Sigma Ventures. Those firms joined in the Series A round, which also included participation from KdT Ventures, Goldcrest Capital, XTX Ventures, Sahsen Ventures, Greentrail Capital and Alexandria Venture Investments.

Shares0
Shares0