As the French startup ecosystem continues to boom — think Mistral, Poolside, and Adaptive — today the Paris-based Bioptimus, with a mission to build the first universal AI foundation model for biology, emerged from stealth following a seed funding round of $35 million. The new open science model will connect the different scales of biology with generative AI — from molecules to cells, tissues and whole organisms.
Bioptimus unites a team of Google DeepMind alumni and Owkin scientists (AI biotech startup Owkin is itself a French unicorn) who will take advantage of AWS compute and Owkin’s data generation capabilities and access to multimodal patient data sourced from leading academic hospitals worldwide. According to a press release, “this all gives the power to create computational representations that establish a strong differentiation against models trained solely on public datasets and a single data modality that are not able to capture the full diversity of biology.”
Bioptimus can move faster than Google, say leaders
In an interview with VentureBeat, Jean-Philippe Vert, co-founder and CEO of Bioptimus, chief R&D Officer of Owkin and former research lead at Google Brain, said as a smaller, independent company, Bioptimus can move faster than Google DeepMind to gain direct access to the data needed to train biology models.
“We have the advantage of being able to more easily and securely collaborate with partners, and have established a level of trust in our work by sharing our AI expertise and making models available to them for research,” he said. “This can be hard for big tech to do. Bioptimus will also leverage some of the strongest sovereignty controls in the market today.”
Rodolphe Jenatton, a former research scientist at Google DeepMind, has also joined the Bioptimus team, telling VentureBeat the Bioptimus work will be released as open source/open science, at a similar level to Mistral‘s model releases. “Transparency and sharing and community will be key elements for us,” he said.
Other models are limited to specific aspects of biology
Currently, AI models are limited to specific aspects of biology, Vert explained. “For example, several companies are starting to build language models for protein sequences,” he said, adding that there are also initiatives to build a foundation model for images of cells.
However, there is no holistic view of the totality of biology: “The good news is that the AI technology is converging very quickly, with some architectures that allow to have all the data contribute together to a unified model,” he explained. “So this is what we want to do. As far as I know that it does not exist yet. But I’m certain that if we didn’t do it, someone else would do it in the near future.”
The biggest bottleneck, he said, is access to data. “It’s very different from training an LLM on text on the web,” he said. And that access, he pointed out, is what Bioptimus has in spades, through its Owkin partnership.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.
