Flagship's machine learning startup Generate bags $370M in latest round with plans for a big hiring spree
As the future of machine learning and AI promises to make major breakthroughs in drug development, a suite of startups is looking to scale their own competing robot brain trusts. An ambitious startup out of Flagship Pioneering’s incubator uncloaked last year with its own spin in that arms race — and now it’s primed and ready for a major expansion in the coming years.
Generate Biomedicines has closed a giant $370 million Series B from founding investor Flagship as well as Alaska Permanent Fund, Altitude Life Science Ventures, ARCH Venture Partners, and funds and accounts advised by T. Rowe Price Associates, the company said.
Generate, which debuted in late 2020, is Flagship’s shot at using machine learning advances to change the game in the development of protein therapeutics, with the company promising much faster development times and discovery success rates. It’s a model being tried elsewhere from companies like Exscientia and Insilico, but Generate sets itself apart with its focus on protein sequencing and the goal to produce de novo biologics.
Since launch, Generate has been hard at work crunching data through its system in an attempt to give it the broadest possible platform for discovery. Now, the company is ready to greatly expand its “wet lab” and biology capabilities, CEO Mike Nally told Endpoints News, and that will spell a huge hiring spree for the company as well as new corporate digs in the Somerville, MA area and another site in Andover.
Currently operating with a workforce of around 80, Generate is planning to make the quantum leap to 500 employees over the course of the next two years while also moving into two new facilities in the coming weeks and months. The Andover site will be used to expand the biotech’s range in structural biology, including a big investment in cryogenic electron microscopy (cryo-EM), a cutting-edge technique used to determine the structure of proteins.
The immediate goal is to move past the work of the past two years, primarily proving Generate’s platform could produce higher binding affinity proteins from precursor reference biologics, into the realm of developing protein drugs from scratch using proprietary data learning, Generate’s chief strategy and innovation officer and co-founder Molly Gibson told Endpoints.
“Since inception, we’ve set out to answer this question whether we can program proteins at the DNA level, and what we’ve done over the past year is take those first insights … really show that at scale and push programs forward,” Gibson said. “We’ve shown that we’ve been able to make significant advances over where we started creating these machine learning algorithms using public data.”
Now, the team said it will be ready to start churning out preclinical candidates by the end of the year with some of those programs expected to hit human trials as early as 2023.
As the clinic moves closer, the Generate team is keeping a close eye on the potential for strategic partnerships given the potential breadth of its platform’s capabilities. Those strategic partnerships, more in the line of substantial R&D team-ups, are a key part of the company’s BD strategy moving forward, Nally said, which the company will have to leverage with its own in-house work.
“The reality is … we can produce far more than we can consume so we want to make sure we complement our expertise in protein engineering with others’ distinct capability in areas like disease-area biology, manufacturing and clinical development,” Nally said. “Ultimately there’s a huge investment across biotech and pharma in understanding target biology so we think if there’s an appropriate task, this technology may be able to address it in a way that very few other technologies can.”