Recursion Pharmaceuticals has agreed to acquire two biotechnology startups developing artificial intelligence tools for drug discovery for a combined $87.5 million, the company said Monday.
The Salt Lake City company will pay $40 million for Cyclica and $47.5 million for Valence, and expects to add roughly 60 employees to its payroll.
“We are moving programs forward, but it's not at the pace that is possible with the right tools,” said Chris Gibson, Recursion’s CEO and co-founder. “We felt like if we built it ourselves, it was going to be two or three years slower than if we found the best teams in the world and brought them in-house.”
Recursion takes an approach to industrial drug discovery that Gibson describes as “full stack,” borrowing language more common to the technology sector than to biotech. In tech circles, full stack developers work on both the front-end — the parts that users see — and the back-end, or the systems that make a website or app functional.
Cyclica and Valence are meant to help Recursion offer a more complete range of drug discovery services. According to Gibson, Recursion started with a list of more than 100 potential acquisition targets and, over the past 18 months, winnowed that down to Cyclica and Valence.
Cyclica, based in Toronto, has two products that enable it to better predict and generate “hits” on drug targets in the development of small molecule medicines. Valence, based in Montreal, focuses on a machine learning technique known as “few-shot learning” — classifying data with a limited number of samples.
Valence is advised by Yoshua Bengio, a noted Canadian computer scientist who also serves as a scientific adviser for Recursion.
In addition to partnering with other companies, Recursion has several drug candidates of its own in clinical testing for rare cancers and other conditions including neurofibromatosis type 2, an inherited disorder where benign tumors grow along nerves. It has more than a dozen programs in discovery stages for cancers, brain disorders, inflammation and other rare diseases.
“We wanted to de-bottleneck the downstream efforts and accelerate our ability to turn biology and early chemistry insights into real molecules that we will be able to, hopefully, test in humans,” Gibson said.
AI-enabled drug discovery is a hot area in biotech, despite some skepticism from industry veterans about how large its impact might really be. Companies like Exscientia, BenevolentAI and Insitro are working to build new models of drug discovery that draw on fields like computational biology and machine learning. Other biotechs, such as Relay Therapeutics, have put machine learning at the heart of their drug development work.
Recursion is among the most well-funded companies in that group. It raised more than $400 million privately before adding another $436 million in an initial public offering two years ago, one of the largest new stock offerings for a biotech since 2018. The company had $473 million in cash and equivalents at the end of March.
Shares of Recursion rose by as much as 10% through early Monday afternoon on news of its acquisitions and earnings. Still, at about $5.50 apiece, those shares have lost more than two thirds of their value since Recursion's Nasdaq debut in 2021.