Bioptimus raises $41 million to develop 'GPT for biology'
A fledgling French artificial intelligence startup has raised $41 million to develop fundamental AI models of biology.
Just as OpenAI's ChatGPT took the world by storm for its super-intelligent generative AI tool capable of conducting natural language conversations in the form of text, Bioptimus has taken this concept but trained its models specifically for downstream biological applications – with its own unique set of Challenges arise as the required clinical training data are often sensitive and not publicly available.
Bioptimus co-founder and CEO Jean-Philippe Vert said the company hopes to deepen understanding of biology by understanding how raw data works, from molecules to whole organisms. This will allow scientists and researchers to model the biological world to “predict disease outcomes” and develop more effective treatments, he said. It's this simulation that makes its technology somewhat similar to ChatGPT's underlying model, Vert said.
“Essentially, it's like GPT for biology – but instead of generating text, we simulate biology,” he said in a statement.
hotbed
France has become a hotbed for AI startups, with generative AI companies across France receiving the majority of funding as of last year. Big funding rounds include Mistral AI's $640 million, H's $220 million and Hugging Face's $235 million – all in the past 18 months.
Biooptimus was founded last year but has already raised $35 million in seed funding. Less than a year after its founding, the company has now raised a total of $76 million in funding, a testament not only to the current AI hype but also to the backgrounds of Bioptimus' six co-founders. For example, Chief Technology Officer (CTO) Rodolphe Jenatton was a senior research scientist at Amazon and Google. Vert, meanwhile, is not only co-founder and CEO of Bioptimus, but also chief research and development officer of French unicorn Owkin, another AI biotech startup whose backers include GV.
This dual role hints at the origins of Biooptimus. Owkin leverages artificial intelligence and machine learning to accelerate drug discovery and has developed extensive partnerships with top biopharmaceutical companies. As part of this work, Owkin has also amassed a large amount of multimodal patient data, which Biooptimus is using to train its base model.
Rather than creating a tangential unit within Owkin that focuses on the base model, it makes more sense to create a separate entity.
“Building biology (basic models) is not part of Owkin's roadmap, but Owkin supports and is keen to work with companies like Biooptimus,” Vert explained in an interview with TechCrunch last year. “Training very large scale (base models) requires significant resources in terms of data volume, computing power and breadth of data patterns that are more easily unlocked as specific entities. As a 'pure player' in the base model, Biooptimus is better suited to do this ”
Over the next few months, Biooptimus launched H-Optimus-0, an open source base model for pathology that was trained on millions of images to aid research into diseases such as cancer and diagnosis. However, with the new $41 million in the bank, the company will enhance its AI platform with more diverse data sources covering a wider range of therapeutic areas, while also looking to form further partnerships with the pharmaceutical and biotech sectors relation.
As part of this, the company is preparing to release a new multimodal base model later this year that spans the entire biological spectrum and could advance developments in areas such as medical, biotech and even cosmetics.
“Beyond pharmaceuticals, this model will open up endless possibilities for many other industries, driving biological discovery in ways we are only beginning to imagine,” said Witt.
Bioptimus’ latest cash injection was led by U.S. venture capital firm Cathay Innovation, with participation from Sofinnova Partners, Bpifrance, Andera Partners, Hitachi Ventures, Boom Capital Ventures, Pomifer Capital, Sunrise and multiple angel investors.