Iambic Therapeutics, Inc.

As he was testing new antivirals to halt a future pandemic, John Chodera encountered an unlikely set of new problems that ultimately helped lead to his new job. The computational chemist won a $68 million grant in 2022 from the National Institutes of Health to help lead a five-year effort to develop new antivirals. That work ended prematurely when Congress decided in 2023 to effectively abandon that initiative. “We found out a five-year program to discover drugs was really going to be a three-year halfway bridge to nowhere,” Chodera said in an interview. “If you give up halfway across the river, you don’t get a useful bridge with half the capacity. You don’t get anything useful at all, unfortunately.” After a decade-plus of running a computational biology lab at Memorial Sloan Kettering, Chodera is now venturing out on his own as a first-time CEO. The antiviral experience wasn’t just a lesson in whimsical pandemic funding, but it brought Chodera closer to the process of discovering drugs. The 47-year-old had previously developed software to simulate molecules, most notably OpenMM , but the drug-hunting effort showed how today’s tools — even AI-powered ones — lacked the accuracy or scalability to do what’s needed in finding, testing and making drugs. Chodera is now CEO and president of Achira, a biotech startup incorporated last summer that launched Friday with a $33 million seed round. Achira is aiming to build the next generation of these computational tools for drug discovery, which Chodera calls “atomistic foundation simulation models.” That’s a complicated way of saying his team is blending AI- and physics-based methods together to try to improve drug discovery. Chodera’s co-founders are Theofanis Karaletsos, head of AI at the Chan Zuckerberg Institute and formerly an AI leader at the biotech insitro, and Zavain Dar, a co-founder and managing partner of Dimension. Dimension led the seed round, which exceeded an initial $15 million target and included Amplify Partners, Compound, and NVentures, Nvidia’s VC arm. Dar said Dimension spoke with over 50 biotechs on the theme of combining machine learning and molecular dynamics, ultimately deciding to start a new company with Chodera and Karaletsos. Those two have known each other for about a decade, and have long discussed the ideas that turned into Achira. “We want to turn drug discovery into 90% compute and 10% experiment,” Karaletsos said. “This is the domain where we can do it based on first principles.” Achira’s view is that neither physics-based methods nor machine-learning models are enough on their own in drug discovery. ML models still don’t have enough drug discovery data, Chodera said. And physics methods, like molecular dynamic simulations that power Schrödinger’s ubiquitous software, are powerful but limited. Those calculations need tons of compute power to show molecules wriggling, moving and interacting with each other. Scientists face a trade-off in how many molecules and how long those simulations can be. Chodera, who previously served on Schrödinger’s scientific advisory board, said the computational framework known as force fields that power those simulations has effectively gone unchanged for over four decades. In the meantime, compute power has grown exponentially. “This stuff has not changed fundamentally in 45 years, which is crazy because those models were built for a PDP-11, a late 1970s-era computer,” he said. “The cost per FLOP [floating-point operation, a cost-effectiveness metric for compute power] is now 160 billion times less than it was at that time, and we still are using the same functional form.” Achira is developing a type of AI model called neural network potentials that rely on the physical laws of nature to simulate molecules. Achira plans to then generate synthetic data by using these models, making large biological datasets that are in short supply today. “The foundation simulation models that we are building are going to be extremely general,” Chodera said, meaning they could subsequently be trained for a range of tasks like predicting a molecule’s binding affinities, selectivity or stability. The 10-employee company has hired several notable scientists, including CTO Daniel Smith, a co-founder of Iambic Therapeutics and most recently head of computation at a Flagship Pioneering startup called Abiologics , and a mix of computational biologists, theoretical chemists, biophysicists and chemical engineers. Achira expects to grow around the hubs of AI talent in San Francisco and New York. Achira’s leaders declined to specify what business model it may pursue, but won’t start by building its own pipeline of drugs. Chodera said he expects to have “some fun stuff to share later this year” on its first generation of its simulation models.

SAN DIEGO--( BUSINESS WIRE )--Iambic Therapeutics, a clinical-stage biotechnology company developing novel therapeutics using its unique AI-driven discovery platform, today announced the appointment of Peter Olson, Ph.D. to the role of chief scientific officer. Dr. Olson, who was most recently vice president, research at Mirati Therapeutics, a Bristol Myers Squibb (BMS) company, will join Iambic’s executive leadership team. A specialist in cancer biology and genetics, Dr. Olson will lead a team focused on driving small molecule programs from target identification through to clinical proof-of-concept. At Mirati, he led in vitro and in vivo pharmacology and translational research across its portfolio, which included pre-clinical efforts for KRAZATI™ (adagrasib), a KRAS G12C inhibitor approved by the US Food and Drug Administration (FDA) in 2022. “We couldn’t be more excited to welcome Pete Olson to the Iambic team,” said Tom Miller, PhD, Iambic’s chief executive officer. “He brings extraordinary drug discovery and development leadership experience as we advance and expand our pipeline of clinical candidates for oncology and other disease areas.” Iambic is developing an internal pipeline of candidates, discovered using its platform that integrates AI models for protein structure prediction and wholistic drug design with high-throughput chemistry and biology experimentation. The Company’s pipeline currently includes IAM1363, a highly selective, brain penetrant small molecule inhibitor of both wild-type and oncogenic HER2 mutants currently in a Phase 1/1b study, as well as a potential first-in-class selective dual CDK2/4 inhibitor for multiple cancer indications, an allosteric inhibitor for KIF18A, and additional new programs. “Iambic’s AI-driven discovery platform enables the rapid identification of promising drug molecules and provides predictive insights into clinical properties of potential medicines,” said Dr. Olson. “I am excited to join the Iambic team to further advance their promising pipeline as we work to deliver innovative breakthroughs for patients.” Dr. Olson has over twenty-five years of basic, pre-clinical and translational cancer biology experience in academia and industry and is the author of close to 50 peer-reviewed papers. Prior to Mirati he was a senior principal scientist in Pfizer’s Oncology Research Unit, where he was the translational project leader for OGSIVEO™, the first FDA approved treatment for progressive desmoid tumors. Before Pfizer, he was a postdoc in the Hanahan Laboratory at the University of California, San Francisco. He received his Ph.D. in Biology from the University of California, San Diego and a B.S. in Biochemistry from the University of California, Davis. About Iambic’s AI-Driven Discovery Platform The Iambic AI-driven platform was created to address the most challenging design problems in drug discovery, leveraging technology innovations such as Enchant, a multi-modal transformer model that predicts clinical outcomes from the earliest stages of discovery, and NeuralPLexer, the best-in-class predictor of protein and protein-ligand structures. The integration of physics principles into the platform’s AI architectures improves data efficiency and allows molecular models to venture widely across the space of possible chemical structures. The platform enables identification of novel chemical modalities for engaging difficult-to-address biological targets, discovery of defined product profiles that optimize therapeutic window, and multi-parameter optimization for highly differentiated development candidates. Through close integration of AI-generated molecular designs with automated experimental execution, Iambic completes design-make-test cycles on a weekly cadence. About Iambic Therapeutics Founded in 2019 and headquartered in San Diego, California, Iambic Therapeutics is disrupting the therapeutics landscape with its unique AI-driven drug-discovery platform. Iambic has assembled a world-class team that unites pioneering AI experts and experienced drug hunters with strong track records of success in delivering clinically validated therapeutics. The Iambic platform has been demonstrated to deliver high-quality, differentiated therapeutics to clinical stage with unprecedented speed and across multiple target classes and mechanisms of action. The Iambic team is advancing an internal pipeline of clinical assets to address urgent unmet patient needs. Learn more about the Iambic team, platform, and pipeline at iambic.ai .

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.