The AI-focused biotech Generate:Biomedicines announced another pharma partnership Tuesday, agreeing to work on several protein therapies with the Swiss drugmaking giant Novartis.
Novartis will pay $50 million in cash and put in a $15 million equity investment while lining up possible milestone payments of over $1 billion. The two didn’t disclose specifics like disease areas or the number of targets. Similar to an ongoing partnership with Amgen, the Generate-Novartis collaboration will focus on a range of protein-based therapeutics, which could include antibodies, peptides, enzymes, or cytokines.
The deal is just the latest example of pharma’s growing interest in partnering with AI-focused biotechs. Novartis and Eli Lilly both announced research deals with Alphabet’s Isomorphic Labs at the start of 2024. This month, the California startup Genesis Therapeutics announced its third pharma partnership in four years with Gilead paying $35 million upfront, and Superluminal Medicines closed a $120 million Series A that included Eli Lilly among its backers.
For Generate, Merck veteran Mike Nally has led a period of rapid growth since joining as CEO in 2021, including raising over $600 million and putting its first drugs into the clinic. Generate’s pipeline is now approaching 20 different programs, Nally said, and he expects to file three to six IND applications over the next 12 to 18 months. The Novartis deal will help the 320-employee biotech apply its AI-based technologies to more drug programs, he told Endpoints News.
“We have a technology that is remarkably scalable in terms of molecular design capabilities that allows us to do more than we can ever prosecute on our own,” he said.
The deal will further pad Generate’s already-lined coffers. Counting the Novartis deal, Generate now has over $350 million in cash, Nally said, which gives the company flexibility when weighing financing paths like a potential IPO.
“If we believe that going public is the best path, we will pursue it,” he said. “But at the same time, we don’t think we need to go public to be successful. We have ample cash today and 25 different institutional investors that have really been great and supportive of us in the private sector.”
Upcoming readouts
Generate turned heads last November when it published on using its AI model called Chroma for de novo protein design. But the biotech’s lead programs, which are nearing early clinical data, are not de novo efforts. Instead, the company has used AI to help tackle the multi-parameter optimization challenge that is drug development, Nally said, in particular by boosting the binding affinity of a drug to its target while maintaining other key attributes like manufacturability.
It already has two programs in the clinic: a Covid-19 antibody and an anti-TSLP antibody.
For its Covid program, Generate will present interim Phase 1 data in October at the IDWeek conference in Los Angeles. Nally said they plan to pause that antibody’s development after the Phase 1 is completed, citing the uncertainty of the Covid market against the need to pay for and run large trials.
The company is also on track to have initial data later this year from an ascending single-dose study of its TSLP program, GB-0895, in mild-to-moderate asthmatics. Generate expects six-month clinical data late in the first quarter or early in the second quarter of 2025, Nally said.
The TSLP drug will be an early test of whether Generate’s AI approach to molecular design can translate to clinical benefit. Its lead programs are targeting well-studied proteins that already have approved treatment options, such as Amgen and AstraZeneca’s Tezspire for TSLP or Eli Lilly’s Ebglyss for IL-13.
Nally said its TSLP antibody showed a 20-fold improvement in binding compared to Tezspire, which he hopes will translate to less frequent dosing. Tezspire is a monthly injection, and Nally said he’s encouraged GB-0895 could be dosed every six months.
The same pitch is behind several of Generate’s next antibodies, targeting IL-13, OX40L and TL1A. Nally said all three have “clear line of sight” to the clinic, with the IL-13 antibody set to enter the clinic early next year. These drugs have shown 20- to 40-fold improvements in the ability to bind to their targeted proteins, he added.
Their internal programs are taking 18 to 24 months to reach the clinic, far quicker than an industry average of three to five years, Nally said.
“More importantly than the time, what we really care most about is quality,” he said. “Are we getting different, better quality answers? We’re seeing that now with the first few programs we’re putting forth.”