Latent Labs, a pioneering startup founded by Simon Kohl, a former Google DeepMind scientist, has recently emerged with an impressive $50 million in funding. This ambitious venture aims to “make biology programmable” by developing artificial intelligence foundation models designed to generate and optimize proteins. The company, which was incorporated in London in mid-2023, endeavors to transform the traditional drug discovery process by collaborating with biotech and pharmaceutical companies, thereby expediting and de-risking early-stage research and development.
Simon Kohl departed from DeepMind at the end of 2022 to establish Latent Labs, bringing together a team of approximately 20 employees spread across two sites in London and San Francisco. The startup operates a wet lab in San Francisco alongside a computational protein design team. Unlike other firms, Latent Labs does not focus on developing its own therapeutic candidates; rather, it enables researchers to "computationally create" novel therapeutic molecules from scratch. The technology driving this innovation is rooted in the work of DeepMind's AlphaFold2, which has predicted the shape of around 200 million protein structures.
“There have been some very interesting seeds planted, [for example] with AlphaFold and some other early generative models from other groups,” said Simon Kohl.
Latent Labs aspires to overhaul the current drug-discovery process, which traditionally relies on numerous experiments and iterations. By making biology programmable, the company aims to reduce the need for wet labs in the future. Kohl elaborates on this vision:
“Our mission is to make biology programmable, really bringing biology into the computational realm, where the reliance on biological, wet lab experiments will be reduced over time,” said Simon Kohl.
The company has secured investments from notable backers, including Radical Ventures, Sofinnova Partners, and Google's chief scientist Jeff Dean. With a leaner setup compared to DeepMind, Latent Labs emphasizes building frontier models for protein design. This approach aligns with its sister company Isomorphic Labs' mission to apply AI research in transforming drug discovery.
Latent Labs' strategic partnerships with biotech and pharmaceutical companies will allow its models to be tested in real-world scenarios. This collaboration will provide essential feedback to ensure the models' progression aligns with the company's objectives.
“This enables us to test our models in the real world and get the feedback that we need to understand whether our models are progressing the way we want,” Kohl explained.
Kohl acknowledges the financial challenges associated with large-scale computations, emphasizing the importance of efficient resource allocation.
“Compute is a big cost for us as well — we’re building fairly large models I think it’s fair to say, and that requires a lot of GPU compute,” stated Simon Kohl.
As Latent Labs continues to innovate in AI-driven protein design, its ultimate goal is to create a paradigm where hypotheses about drug targets can seamlessly translate into viable protein drugs with all desired properties incorporated.
“Imagine a world where someone comes with a hypothesis on what drug target to go after for a particular disease, and our models could, in a ‘push-button’ way, make a protein drug that comes with all of the desired properties baked in,” envisioned Simon Kohl.
Leave a Reply