Behind the Term Sheet: Bioptimus’ $41M Series A
The GPT of Biology Raising the Bar of AI-Driven Scientific Research
We’re starting to get accustomed to AI taking us to uncharted territory, transforming virtually every aspect of industries while reshaping how businesses and individuals access and process information. This has been in part achieved through the rise of foundation models, with major players emerging across the world from OpenAI in the US, Mistral in France to China’s recently released DeepSeek.
While foundation models have been around for several years, their impact has tremendously increased. Initially, mainstream adoption was predominantly centered around more “basic” tasks such as text or image generation and information gathering through simple prompts — relying on public & internet data to boost productivity and efficiency. But this is only the tip of the iceberg. As AI models continue to evolve — taking into account new multimodalities, becoming context-aware and capable of reasoning across complex datasets — their influence is expanding into more advanced applications, knowledge-driven industries and even the unknown.
For instance, imagine an AI foundation model that had a holistic understanding of complex biology — connecting the dots between modalities and scales — from molecules to cells, tissues, and whole organisms? This would unleash untold biotechnological research and innovation. And this is precisely where Bioptimus comes into play.
Bioptimus — developing “the GPT of biology” — recently announced its $41M Series B led by us at Cathay Innovation. The round comes as a natural continuation of our support in its seed round less than a year ago, not to mention the company is incubated by Owkin, the French AI biotech leader Cathay Innovation has backed as early as 2017. This is a testament to the trust in Bioptimus’ potential to harness state-of-the-art AI technologies to unlock unprecedented possibilities for a wide variety of industries, from early drug design processes to patient diagnostics.
FOUNDATIONAL AI x BIOLOGY
Over €100B is spent every year on pharmaceutical R&D, which faces declining ROIs and shrinking budgets. This puts pressure on pharma companies to find more cost-effective, time-efficient, and data-driven solutions to sustain drug discovery efforts and develop new life-changing medicines and treatments.
While there are many factors at play — one of the most glaring is the immense complexity of biology, where even small molecular changes can trigger unforeseen effects across interconnected systems, and the ongoing human pursuit to understand it.
There have been significant advancements in applying AI to drug discovery and medical research, but AI is far from being a miracle, cure-all solution. It’s only as good as the data it’s trained on — and in the case of biology, that data is historically fragmented. While medical and patient data are highly regulated, research has also long been conducted in silos, focusing on isolated components such as DNA, proteins, cells, or tissues rather than capturing the full complexity of biological systems.
Several AI models have already made breakthroughs in specific areas of biology, like DeepMind’s AlphaFold, Meta’s ESMFold (protein sequencing), or Alphabet’s Isomorphic Labs (AI-driven drug design). However, these models remain specialized — they operate within narrow problem spaces, focusing on a single biological level and modality and are limited in size and complexity.
To truly unlock AI’s potential in biology, there is a need for universal models that integrate data and insights across molecules, cells, tissues, and entire organisms — allowing for deeper understanding, more accurate predictions, and ultimately, breakthroughs in medicine and biotechnology.
ENTER BIOPTIMUS: BREAKING FREE FROM SILOS IN BIOLOGY
Bioptimus is redefining biological research by addressing the siloed nature of biological understanding. With the mission of building the world’s first AI foundation model for biology, its revolutionary approach integrates data across multiple scales (molecules, cells, tissues, organisms) and modalities (imaging, proteomics, etc.) — for a holistic view of biology as it exists in reality.
“What we are building is not just a technological breakthrough; it’s a transformative tool for understanding biology in its full complexity. By learning how biology works directly from raw data across scales, from molecules to whole organisms, our model will empower researchers in the pharmaceutical industry to simulate complex biology, predict disease outcomes and response to treatment, and design therapies with unprecedented precision. Beyond pharmaceuticals, this model will unlock limitless possibilities across many other industries, driving biological discoveries in ways we are only beginning to imagine. Essentially, it’s like the GPT of biology — but instead of generating text, we’re simulating biology.”
Jean-Philippe Vert, co-founder and CEO of Bioptimus
What Excites Us Most About Bioptimus:
- The Handpicked Team of Seasoned Experts: The Bioptimus team’s credentials speak for themselves, with a founding team of big tech and biopharma veterans hailing from Google, Amazon, and Cathay Innovation-backed Owkin, the first Franco-American unicorn in AI for medical research. At the helm of Bioptimus is co-founder and CEO Jean-Philippe Vert, also Chief R&D Officer at Owkin and former head of Google Brain in Paris, where he led projects on AI for biology, including protein LLMs and FMs for single-cell omics. He’s surrounded by a world-class team of 17 (and counting) scientists, researchers, and engineers (many of whom are Google DeepMind alumni), with some of the best credentials in ML academia has to offer, from institutions like Polytechnique, ENS, Berkeley, MIT, and on.
- The Data Edge & Owkin Connection: In the world of FMs, (public + proprietary) data is king. While any model can access and train from public data, it’s the non-public data sets that serve as a huge value add and differentiator. This is where the Owkin connection makes a difference– Bioptimus complements publicly available data with private data sets via partnerships with research labs as well as exclusive access to Owkin’s MOSAIC data (the largest multimodal atlas of cancer with spatial biology to connect molecules and tissues) and larger-scale network.
- The Traction: Bioptimus has uncapped potential to tackle a combined market of GenAI and biomedical research with a foundation model that will be used by all new downstream tools. And just as LLMs did with natural language, Bioptimus’ model will learn to speak biology instead of human language. In less than a year, the Bioptimus team has already made some remarkable progress, successfully releasing its first model H-Optimus-0 — now the world’s largest foundation model for pathology with an unparalleled ability to predict gene expression from morphology and to subtype ovarian cancer with exceptional accuracy. Of note, the model was trained in less than 3 months and has already outperformed all other pathology models in independent benchmarks, including assessments from Harvard Medical School’s HEST program and the University of Leeds.
What’s next for Bioptimus? The team is gearing up for the launch of its breakthrough new multi-scale, multi-modal foundation model for biology this year — allowing for the simulation of biology at unprecedented scale and dimension. In the meantime, the company will continue to focus on building out the team and integrating even more diverse data sources and therapeutic areas, cementing new strategic partnerships with pharma and biotech and expanding critical datasets to further refine and validate its models.
PARTING THOUGHTS — NEXT GEN RESEARCH: PRECISE, UNIVERSAL, GROUNDBREAKING
At Cathay Innovation, we follow a simple M.O.: investing in companies that trigger positive, meaningful and durable change. Digital Health has been and will continue to be near to our hearts. Backing Bioptimus was a no-brainer from the start and a logical extension of our roster of AI-powered portfolio companies in the space from Owkin to Nabla, AQEMIA, Entalpic and many more.
For years, we have opted for an investment tactic that has proven fruitful, with their success only making us even more confident in the importance of collaboration between our corporate ecosystem — think Sanofi — and emerging game-changers. This is just the beginning of an AI revolution that will truly transform the conduction and evolution of medical research and life sciences as a whole, and we are excited to witness it firsthand.
For more insights on the world of AI, digital health and startups across our key sectors, check out our previous posts:
