CEO: Chris Gibson
Based: Salt Lake City, Utah
Recursion Pharmaceuticals has raised $73 million over a 12-month period, from October 2016 to 2017, with its series B taking the lion’s share of a chunky $60 million as it looks to step up its artificial intelligence-enabled drug discovery activities and pivot away from the relatively simple job of repurposing meds. It boasts of pacts with the likes of Sanofi and Takeda, which has given the young startup confidence in its approach.
What makes Recursion fierce
Salt Lake City-based Recursion is only four years old but already has big plans: CEO Chris Gibson has boldly stated that he wants the company to have discovered 100 treatments by 2025, underpinned by speed.
Gibson explains: “Our data scientists are already using artificial intelligence [AI] methods to query maps of cell biology and pharmacology to make predictions about the properties of potential drugs and their effects on hundreds of diseases. Using a combination of these approaches and automation, we’re able to shave months to years off of the drug development process already. As we bring more data into our models, our goal is to deploy our AI methods to accelerate nearly every stage in the drug development pipeline.”
It is slated to file for its first IND next year, becoming a fully fledged R&D vehicle after an initial focus on repurposing molecules.
“Working to repurpose drugs in rare genetic diseases provided us a unique opportunity to potentially discover drugs for thousands of patients across many diseases while also capturing a wealth of data from genetic manipulation of human cells and well annotated drugs from the repurposing toolbox,” Gibson says.
“As we demonstrated to ourselves, significant learning in that constrained space, the obvious and always planned expansion has been into new biology (inflammation, immuno-oncology, aging, and infectious disease) and to expand into new chemical entities and target discovery.”
He says that the data his company is collecting and the computational tools it’s building “are about much more than building a typical drug discovery pipeline that you’d expect for a biotech or even a large pharma.”
The company wants to use cutting-edge computation to build in efficiencies at each step, in ways that first enable “radical scale of the pipeline,” but ultimately drive singularity-level changes in the way it selects the right drug for the right patient. “We’re driving toward realizing the promise and power of truly precision therapeutics, and we believe the way to get there is by starting with massive quantities of highly controlled and curated biological data.”
What to look for
Those 100 discoveries by the middle of next decade for one thing, and for another, building on the promises of computational biology, from Recursion and the wider industry. Gibson says that these promises “have been made and broken before, both in the 1980s (new drugs will be designed by computers!) and in the 2000s (gene sequencing will provide the blueprint for life and make sense of it all)!”
He says that while these technologies have contributed to many improvements and myriad successes, they haven’t quite been the holy grail they aspired to be. “This time,” he says, “things might be different,” because of three main areas: technologies that have emerged to allow companies to collect significantly more data than ever before (Recursion can generate more than 20 terabytes of image-based data each week); second, it’s now much less expensive to analyze data than ever before; and third, advances in analyses like deep learning are likely to change the game in their ability to draw more humanlike inferences from massive data sets that are millions of times too big for a human to digest.
“So, while it is clear that we are nowhere near utilizing the full utility of ‘AI’ or ‘machine learning’ in the biopharma industry, we are poised to see gains we’ve never before experienced over the coming years,” Gibson explains.
“While promising, everyone is holding their breath for when the first drugs discovered by AI methods demonstrate efficacy in humans, which is when we’ll start to see much more widespread adoption (hint: we’re a very small number of years away from this now).”