BioPharmaTrend.com

The “4th wave” of computational drug design is here, according to a new report!  🤖 🧬 I’ve come across a really well-put report on the evolving state of artificial intelligence (AI) in drug discovery in 2024, authored by Chris Bradbury from Stanford School of Business. The report delivers a crucial insights into why many existing AI companies failed to deliver on a decade-old hype in drug discovery, and why we are likely on the cusp of a new era in computational drug design with companie slike Nabla Bio, NOETIK, Iambic Therapeutics and some others. The report breaks down the AI progress in pharma/biotech into four "waves" based on the tech stack and data generation approaches of the comapnies, and their respective business models. Also, my five cents about all this and some conclusions. Kudos to Dylan Reid for pointing my attention to this wonderful resource! I enjoyed every slide of the report and I am a bit envy it is not my work 😊 #foundationmodles #genAI #biotech Image source: my 2023 keynote speach at Deloitte center in Brussels, about foundation models in biology.

Today's newsletter is out! 🔷 Mapping the Epicenters of Pharma Innovation: Brian K. Buntz and Maryam Daneshpour analyze global #pharma innovation hubs using a dual-weighted metric of patent activity and market capitalization. 🔷 Foundation Model for Precision Radiology: French startup Raidium, founded by Paul Herent, M.D., Msc and Pierre Manceron, is building a multimodal foundation model for medical imaging. Trained on over a billion real-world images, their AI aims to assist radiologists with tasks like metastatic cancer analysis and biomarker discovery. Raidium raised €16M in seed funding, backed by the EIC, and presented at the Radiological Society of North America (RSNA) Congress in Chicago this week. 🔷 Injectable Immune Organs Against Ovarian Cancer: Dr. girija goyal and her team at the Wyss Institute at Harvard University are developing "iNodes," injectable lymphoid organs designed to stimulate immune defenses against #ovarian #cancer by transforming cold tumors into hot ones. The project is now backed by an Advanced Research Projects Agency for Health (ARPA-H) Sprint for Women’s Health award. 🔷 Insilico Medicine's FDA Milestone: Insilico Medicine announced FDA IND clearance for its 10th AI-designed drug, ISM5939, targeting ENPP1 for solid tumors. The candidate was designed in just three months using their Chemistry42 platform. Read the full stories in the comments.

In today's newsletter we take a look at: 🔷 The first draft of the Human Cell Atlas (HCA) is here, offering comprehensive maps of over 62 million cells from 9,100 donors. This collaborative effort involves 3,600 researchers across 102 countries, setting the stage for breakthroughs in personalized medicine, disease modeling, and drug discovery. From multiomics to spatial transcriptomics, the HCA integrates cutting-edge techniques to catalog human cells in unprecedented detail. 🔷 Using Joint Atomic Modeling (JAM), Nabla Bio is generating antibodies from scratch, bypassing the need for known binders. This computational approach has already shown success against challenging targets like Claudin-4 and GPCRs, marking a leap forward in therapeutic antibody development. 🔷 Recursion and Exscientia have finalized their merger, combining Recursion OS with Exscientia’s AI-driven molecular design expertise. With an AI world model and wet lab experiments working in a complementary loop, Recursion plans to develop virtual cells to run clinical trials at scale. 🔷 Sanofi, Formation Bio, and OpenAI introduce Muse, an AI platform designed to overcome one of the biggest challenges in clinical trials—patient recruitment. By identifying optimal patient profiles and automating outreach, Muse has the potential to cut recruitment timelines from months to minutes, accelerating trial progress and ensuring broader representation. Read the full stories in the comments!

🌟 How soon will we realistically create a robust “virtual cell” model, allowing simulations of cellular processes to aid drug discovery? 🌍 How to adapt to a rapidly changing clinical trial environment? 🤝How to win more positive public perception of the pharmaceutical industry? I am still reflecting on the learnings about these and many other topics from the recent extraordinary FT Live 𝐏𝐡𝐚𝐫𝐦𝐚 & 𝐁𝐢𝐨𝐭𝐞𝐜𝐡 𝐒𝐮𝐦𝐦𝐢𝐭. There, I had the privilege of meeting some of the most influential figures in the industry, including Seth Ettenberg, President and CEO, BlueRock Therapeutics; Sebastian Guth, COO, Bayer; Kabir Nath, CEO, Compass Pathways; Lord Patrick Vallance, Minister of State, Department for Science, Innovation and Technology; Kasim Kutay, CEO, Novo Holdings; Kris Sterkens, Company Group Chairman, Johnson & Johnson Innovative Medicine (EMEA). But I was especially impressed and inspired by meeting Emma Walmsley, CEO, GSK, who in my opinion is the greatest role model for aspiring female entrepreneurs in #biotech and drug discovery! Among many talks, I recall an inspiring conversation between Demis Hassabis, CEO of Google DeepMind and this year’s Nobel Prize winner and a bright and energetic journalist Andrew Dunn, from Endpoints News. Kudos to Financial Times, and Endpoints News for organizing this phenomenal event and for inviting the BioPharmaTrend.com team as media attendees! And for all whom I met there, please reach out and let’s keep the conversation going! #PharmaLeadership #FTPharma #biotech #LifeSciences Where Tech Meets Bio (Substack Newsletter) #VirtualCell

📩 In this week’s newsletter: • AlphaFold3 Goes Open Source: Google DeepMind and Isomorphic Labs release AlphaFold3, advancing protein structure prediction with a 20% improvement in accuracy for protein-ligand interactions. Now publicly available on GitHub, it promises enhanced speed, broader scope, and new research possibilities—albeit commercial restrictions remain. • Generative AI Tackles IPF: Insilico Medicine shares positive Phase IIa results for ISM001-055, a generative AI-designed drug for idiopathic pulmonary fibrosis. The trial demonstrated lung function improvements and quality of life gains for patients. • Rethinking Gene Delivery: EvoCapsid emerges as a hackathon-to-startup project, leveraging foundational AI tools like AlphaFold and ESM3 to design stealthy, immune-evasive vectors composed of synthetic components that resemble human proteins for gene therapy. • OpenPhenom by Recursion: A new foundation model for microscopy data analysis, outperforming CellProfiler, is now available for non-commercial use. Meanwhile, Recursion continues to innovate with proprietary models like Phenom-1 and Phenom-2. 🔗 Read the full stories in the comments!

A new AI foundation model in bio is claimed to set the new "gold standard" for microscopy data analysis, outperforming CellProfiler. Now, Utah-based Recursion launched a new model, OpenPhenom, which is a non-commercial, publicly available foundation model built on microscopy data that the company believes has set the new "gold standard" outperforming CellProfiler. Large-scale cell microscopy assays are crucial for discovering new biological processes, drug candidates, and potential targets. Traditional software for analyzing these assays, however, requires constant fine-tuning for each new experiment and can easily break down if assay conditions change. As massive datasets like RxRx3 and JUMP-CP grow, these outdated tools struggle to keep up with the data demands. OpenPhenom-S/16, designed by Recursion, offers a powerful alternative for non-commercial researchers. Company says it can replace old workflows with an adaptable, ready-to-use model that consistently beats traditional analysis methods, with no extra tuning or training needed. For example, OpenPhenom-S/16 significantly enhances recall of known biological relationships in the JUMP-CP cpg0016 dataset, as cross-verified with the StringDB database. Notably, Recursion also has models for commercial use for its own pipeline and partnership programs. Recursion has featured the performance of two generations of proprietary models trained on Recursion’s internal microscopy, Phenom-1 (introduced at CVPR 2024) and Phenom-2 (introduced at NeurIPS FM4S 2024) -- also see image below. Recursion proprietary models are stronger than OpenPhenom (surprise!) but that does not diminish the value the public model has for the drug discovery community. Interestingly, just yesterday we covered another foundation model by Recursion -- MolE, for molecular property prediction (link in comments). Undoubtedly, foundation models are on the rise in biotech! I previously compiled a list of 19 companies pioneering AI foundation models in pharma and biotech (link in comments), and two new arrivals by Recursion, and some other new foundation models on the market, prompted me to update the article (follow me to stay tuned!). Image credit: Recursion

BREAKING: AI-driven drug shows promising Phase IIa results in treating fatal lung disease! Insilico Medicine just announced breakthrough results from its Phase IIa clinical trial for ISM001-055, an AI-designed drug candidate for idiopathic pulmonary fibrosis (IPF). This novel treatment could transform options for IPF, a disease with very few effective therapies available today. 📌 Why This Matters IPF is a chronic, progressive lung disease that affects nearly 5 million people worldwide, with a median survival of only 3-4 years. Current treatments help slow disease progression but can’t stop or reverse it. Insilico’s ISM001-055 was developed using cutting-edge generative AI, targeting TNIK (Traf2- and NCK-interacting kinase), a key driver of lung fibrosis, to offer a potentially disease-modifying treatment. 🔬 Key Findings from Phase IIa ✔️ Trial Design: Double-blind, placebo-controlled study with 71 IPF patients across 21 sites in China. Patients were given one of three doses or a placebo over 12 weeks. ✔️ Lung Function Improvements: Patients on the highest dose (60 mg) showed a mean increase in forced vital capacity (FVC) of 98.4 mL, contrasting sharply with a mean decline of -62.3 mL in the placebo group. ✔️ Quality of Life Boost: The highest dose group also experienced a significant 2-point improvement on the Leicester Cough Questionnaire (LCQ), indicating better symptom control and quality of life. ✔️ Safety: ISM001-055 was well-tolerated, with only mild side effects reported, primarily gastrointestinal issues and minor liver function changes. 🧬 What’s Next? To build on these results, Insilico is now recruiting for a parallel Phase IIa trial in the United States to further validate the efficacy and safety of ISM001-055. If successful, this drug could shift the treatment paradigm for IPF. 🌐 Broader Impact Insilico Medicine’s PharmaAI platform seems to be a productive tool for designing first-in-class treatments faster (and arguably, quite efficiently, judring by ongoing clinical progress). Since 2021, Pharma.AI helped nominate over 20 preclinical candidates across a portfolio of more than 30 assets, with 9 molecules receiving IND approvals. ISM001-055, discovered and designed using @PharmaAI’s Biology42 (PandaOmics) and Chemistry42 modules, is the latest example of how AI can make impact on complex, life-threatening diseases like IPF. An important test for company's abilities will probably be next year when a landmark Phase III trima may launch. Full coverage on BioPharmaTrend.com (link in the comments). Image credit: Insilico Medicine

In this week’s newsletter, we’re sharing some of the first insights from the FT Live Global Pharma and Biotech Summit: • What’s Next for AI in Drug Discovery? Demis Hassabis, CEO of Google DeepMind and Isomorphic Labs, shared future goals, including AlphaFold3’s focus on molecular dynamics and a “virtual cell” model for simulating cellular processes. With partnerships at Novartis and Eli Lilly and Company, Isomorphic Labs is advancing AI-driven research without establishing wet labs. Note: #AlphaFold3 code has just been open-sourced! • Adapting to Change in Clinical Development: The shift toward hybrid outsourcing and decentralized trials is helping pharma contain costs in Phase 3 studies and improve data accuracy in diverse populations. Startups, facing tighter capital, are prioritizing assets with clear revenue potential. • AI Democratizing Drug Discovery: In a separate article, Andrey Doronichev, CEO of OPTIC, discusses how generative AI accelerates and democratizes drug discovery, empowering smaller biotechs and streamlining digital development. 📩 Read the full article in the comments!

We recently sat down with Rui Rui Zhang Xiang from MEDSIR to dive into the practical side of transforming oncology trials. Rui Rui shared how adaptive trial designs and precision medicine approaches are helping address some of the field’s toughest challenges—from the complexities of regional regulations to the use of AI for streamlining trial outcomes. Our conversation also touches on adaptive approaches, such as chemotherapy-free trials for HER2-positive breast cancer, and how partnerships with patient advocacy groups enhance trial alignment with patient needs. To read more, read the full interview in the comment section. Photo: MEDSIR team

Just a heads-up: our CEO, Tanya Bell, is at the FT Live Global Pharma and Biotech Summit, where industry leaders are tackling key topics in drug discovery, clinical trials, and patient engagement. Tanya is here to explore strategic partnerships in the #pharma and #biotech sectors for thought leadership campaigns and joint initiatives to educate public on emerging tech. If you’re around, feel free to reach out! Watch the event live: https://lnkd.in/dcziMy99 #FTPharma