Iambic Therapeutics, Inc.

Iambic Therapeutics has used its AI models to build a pipeline and put its first drug candidate in the clinic earlier this year. Now, the San Diego company is debuting a model with even bolder hopes for what AI can tackle in biotech. On Tuesday, Iambic announced Enchant, a transformer-based model to make clinical predictions. The model is trained on a wide array of preclinical and research data, along with a small amount of clinical data. In a white paper , the biotech said the model exceeded state-of-the-art performance in predicting certain properties like the half-life of a compound in the human body. Its leaders believe the model will excel at making many more predictions that ultimately determine how good an experimental drug will be in humans. “We’ve arrived at a really exciting point with this technology,” Iambic’s co-founder and chief technology officer Fred Manby told Endpoints News . “This demonstration that, yes, you get better at clinical predictions by training on more preclinical data really feels like a moment where something qualitative has changed about the way you can leverage AI in drug discovery.” Iambic’s leaders said Enchant aims to “break the wall” between preclinical and clinical research. The biotech already uses the model’s predictions for properties like half-life or solubility of compounds when selecting preclinical drug candidates. Other early efforts in this space include Google DeepMind’s Tx-LLM , Insilico’s nach0 , and OpenAI’s ChatGPT . Iambic CEO Tom Miller said Enchant’s performance outranks these other models, in a different way than how DeepMind led the competitive space in protein-structure prediction with AlphaFold. He described Enchant’s performance in making clinical predictions as a “zero-to-one situation” for the AI bio field. “The state-of-the-art has nothing to do with DeepMind,” Miller said. “This is a situation where that is just one among a bunch of other methods that aren’t very good and aren’t surpassing the state-of-the-art.” Specifically, Enchant beat the publicly available state-of-the-art performance for predicting the half-life of compounds. On a zero-to-one scale measuring how its predictions match reality, Iambic’s model scored 0.73, showing a moderate-to-strong correlation. The previous record was a 0.58 from a model developed by a research team at a small San Diego biotech called Expert Systems. Iambic started working on Enchant about a year ago, feeding it a wide variety of data, from PubMed and bioRxiv to the Protein Data Bank and Iambic’s own laboratory data. Enchant’s design — a multimodal transformer — mirrors ChatGPT in relying on the AI to make connections across seemingly disparate datasets. Iambic’s leaders believe doing so could better connect lab research on drugs to how they actually perform in the human body — a critical challenge for boosting clinical success rates. Iambic described parts of its performance, like the half-life metric, in a blog post, which was not published in a peer-reviewed journal. Manby said he expects future publications on Enchant and referenced a paper presented in July at a machine learning conference describing an earlier version of the model. While the new white paper provides some examples, it’s hard to tell if those are hand-picked measures of where it shines, or if Enchant is on track to predicting a wide range of clinical properties. Its leaders said they are confident it can beat state-of-the-art performance on many measures beyond half-life. “We’re optimistic we’ll be able to reproduce that across a whole slew of different clinical endpoints,” Manby said.

Terray Therapeutics has raised a $120 million Series B round, aiming to get its first drug candidate into the clinic while advancing in its AI-fueled approach to making molecules. Founding CEO Jacob Berlin exclusively told Endpoints News that the 110-employee startup is generating the “key dataset missing for AI-driven drug discovery” in studying billions of protein-small molecule interactions. Terray has developed tiny chips that can generate more biological data than previously possible. Berlin has spent the past 12 years developing these chips, or microarrays, which are about the size of a nickel and contain millions of wells to hold molecules in place. He left a tenured professorship at City of Hope to start Terray in 2018, which now operates out of a 52,000-square-foot laboratory in the foothills of Los Angeles. The biotech uses these chips to see how tens of millions of molecules bind — or don’t — to a protein of interest. That echoes a common refrain that AI breakthroughs only happen with enough of the right data. And getting those biological data require the right hardware. In the case of the recent Nobel Prize-winning protein model, AlphaFold, it was built off the Protein Data Bank, a painstakingly curated public resource that itself was accelerated by then-new crystallography and sequencing machines. Terray has generated its own proprietary dataset of protein-molecule interactions 50 times larger than publicly available data. The biotech has used that data to train AI models to discover and design small molecules in a new way for a biopharma company, placing the AI’s ideas at the center of its R&D process. The question is if these chips, data and models will ultimately make better drugs. Terray’s Series B will provide multiple years of runway to start finding an answer, which includes aiming to enter the clinic with its lead drug in 2026, Berlin said. Terray is keeping its pipeline under wraps, beyond disclosing a general focus on immunology. Terray is one of several AI bios taking that chance to build its own pipeline. AI-focused startups also working on small molecules include Iambic Therapeutics, Exscientia and Atomwise. Terray CFO and COO Eli Berlin, the brother of the CEO, said the biotech’s hardware has been key in attracting investors in this market. “People recognize the difference in our story,” he said. “That hardware innovation really resonated, because it lets us build these scaled datasets that are precise and iterate rapidly.” Terray’s chips helped design its first molecules, which Jacob Berlin said are often in new chemical space. “Every single molecule we bring forward has a scaffold that has been de novo discovered and optimized at Terray from the ground up,” Berlin said. “That’s totally different than what you see across the ecosystem, where AI models have to stay close to historical sets of data because they don’t have our hardware innovation.” Terray’s process starts by studying a protein target of interest against 80 million molecules, synthesized on nanoscale beads and placed on its chips. After first looking for binding molecules, they then zoom in on the hits to run more screens. As Terray brings compounds into lead optimization or the stage of refining a lead into a drug candidate, AI models suggest what to do next. Humans still make the decisions, Berlin said, but AI increasingly shapes Terray’s experiments. For instance, the company built a diffusion model to generate small molecules. In an August preprint , Terray’s team put their own twist on the diffusion architecture. They require it not only to diffuse out from a starting molecule to generate new compounds but also to diffuse backward to that starting point. That acts like guard rails in restricting the model’s creations, discouraging it from generating radically different molecules. “Computationally, it basically inherently bounds just how wild it gets,” Berlin said. For the molecules at the end of this process, Terray is keeping them under wraps, but Berlin said they have completely new designs that may raise some industry eyebrows. “In every case, med chemists would be surprised. They would say, ‘Huh,’” he said. “I’m surprised by what we find and what we move, so I think other people will be too.” Editor’s note: This story has been corrected in referring to Terray’s chip as microarrays, not microassays.

Artificial intelligence has its Nobel Prize moment. On Tuesday, the Nobel Prize in Physics went to a pair of longtime AI pioneers in Geoffrey Hinton and John Hopfield. Then, on Wednesday, the committee awarded the Nobel Prize in Chemistry to David Baker, Demis Hassabis and John Jumper — three scientists leading the way in using AI to predict and design proteins. Baker, Hassabis and Jumper’s research is behind the creation of some of biotech’s most ambitious startups. That includes Isomorphic Labs, where Hassabis is CEO; and Xaira Therapeutics, which debuted earlier this year with over $1 billion in funding with Baker as a co-founder. The awards reflect the incredibly rapid rise of AI in scientific research — especially the prize going to Hassabis and Jumper, who created AlphaFold, an AI model that predicts protein structures. The Nobel committee sometimes has a reputation for awarding work many years later, waiting to see if a breakthrough has stood the test of time. AlphaFold’s impact on science is still early — and arguably unclear — coming less than four years since the debut of AlphaFold2, a model that quickly predicts the three-dimensional shapes of proteins with unprecedented accuracy. It’s one of the fastest turnarounds in Nobel history in the short timespan from breakthrough to Nobel Prize. But the award “shows the impact of AI on chemistry and medicine and the Nobel Committee clearly understands the impact of AI on everything,” ARCH managing director Bob Nelsen, a key investor in Xaira, wrote in a text. “This is just the beginning.” Hassabis, 48, co-founded DeepMind in 2010, chasing a long-term vision of achieving artificial general intelligence. The project began by training AI models to play board games. Google acquired the startup in 2014. On a flight back from Seoul in 2016, as his team celebrated their AI system beating a world-class player at the board game Go, Hassabis said he realized the technology was now ready to take on bigger challenges. A couple of years later, DeepMind unveiled AlphaFold by entering a long-running competition where research groups attempted to predict the shape and structure of proteins. Determining the structure of a protein has long been a key challenge to figuring out the role of a protein and designing more targeted drugs. At the time, the winning scores had plateaued for years, hovering around 40% accuracy. AlphaFold won with about 60% accuracy. The true breakthrough, though, didn’t come until late 2020, after Hassabis and Jumper rebuilt AlphaFold. The new version — AlphaFold2 — lapped the field and led to claims that DeepMind had solved the protein-folding problem. Less than four years after that moment, Hassabis and Jumper have won the Nobel, as the DeepMind chief is now charging into drug R&D. In an exclusive interview last year with Endpoints News , Hassabis said drug discovery is “the number one thing I always wanted to apply AI to once it was sophisticated enough and powerful enough.” In 2021, he founded Isomorphic Labs, where he remains CEO in addition to leading Google DeepMind. Isomorphic signed partnerships with Novartis and Eli Lilly earlier this year and unveiled AlphaFold 3 in May , predicting far more complex structures than single proteins. AlphaFold 3 can predict the structures and complexes of proteins, DNA and RNA strands, and small molecules among other biomolecules. At a Wednesday press conference after the prize was announced, Hassabis said Isomorphic’s work is “progressing phenomenally well” beyond AlphaFold, exploring areas like designing compounds, predicting binding strength of molecules, and predicting properties like toxicity. “We think there’s huge potential there to revolutionize, actually, the way drug discovery is done and try to shorten it down from almost a decade or more of work, usually, to get a drug all the way through out into the world, to instead of years, maybe months,” Hassabis said. “That could well be possible in the next few years once we continue to develop these technologies further.” Jumper said he’s excited by AlphaFold’s relevance to the earliest stages of research, particularly in identifying biological targets. More than two million researchers have used AlphaFold to date, and about 30,000 scientific papers have cited their use of AlphaFold, he added. “It just really shows that we’re going to start to get good at biology, both on the therapeutics side and the understanding of disease,” Jumper said at the news conference. “The moment I will be almost as excited as this is the Nobel Prize that talks about the work done with AlphaFold to understand this disease.” Baker, for his part, has grown his Seattle laboratory into a powerhouse since founding the Institute for Protein Design in 2012. He began working on a computer model called Rosetta in the late 1990s to predict protein structures. But his lab’s recent success has come from a willingness to embrace cutting-edge AI ideas, like the transformer-based model, after AlphaFold2’s triumph in 2020. His lab has published a torrent of papers over the last few years on creating de novo proteins that look more and more like the leads of potential drug candidates. It has also become a prolific startup creator: Baker has co-founded 21 companies over his career, more than half of those in the past five years. That includes companies at the forefront of AI in biotech like Xaira, Charm Therapeutics, Vilya and Monod Bio. Beyond their corporate ventures, the newly-minted Nobel laureates have had outsized impact in shaping the thinking of the broader AI bio field. Countless startups have launched with the elevator pitch of making “AlphaFold-for-X.” Another slew of young companies, like EvolutionaryScale, Chai Discovery, Iambic Therapeutics and Profluent Bio, have built and use their own protein structure prediction models that compete with DeepMind’s AlphaFold and Baker’s RoseTTAFold. Nobel Prize selections are often supported with a story of how that breakthrough translates to the real world. Recent winners like CRISPR/Cas9 and mRNA, for instance, have produced successful, transformative medicines. In bowing to the AI trend of 2024, Stockholm is placing its own bet that AI’s potential will pay off. Even the newest winners see this as just the start of a longer journey. “I hope that when we look back on AlphaFold, it will just be the first proof point of AI’s incredible potential to accelerate scientific discovery,” Hassabis said.